There’s a lyric from the Oasis song “Wonderwall” that’s been living rent-free in my head every time I read another paywall story: “I said maybe, you’re gonna be the one that saves me.” That’s the emotion publishers brought to paywalls when the model went mainstream in the early 2010s. Not strategy. Emotion. Hope masquerading as a business plan.

Subscribe now

I don’t say that to be cruel. I say it because I was in the room. I watched publishers, smart operators with real businesses, latch onto the paywall narrative the way drowning people grab floating debris. The economics of digital media were already brutal. Print was collapsing. CPMs were a fraction of what the industry needed. And here, finally, was something that felt like agency. We’ll charge for content. We’ll own our audience. We’ll escape the platform trap.

Fifteen years later, we have enough data to render a verdict. And the verdict is complicated, more complicated than the paywall evangelists or skeptics will admit. A handful of publishers built real businesses behind paywalls. Most didn’t. And now, just as the model was finally maturing into something defensible, agentic AI has arrived to stress-test the entire premise.

This is the story of what actually happened, why it happened, and what publishers who are still standing need to understand about what comes next.

Act I: The Promise (2010–2018)

Salvation Narrative

The Wall Street Journal had been running a hard paywall since 1996, a fact that sounds prescient in retrospect but was largely irrelevant to the broader industry at the time. The WSJ served a professional audience whose employers often footed the bill. It was a B2B subscription product wearing a B2C costume. Everyone acknowledged it worked. Everyone assumed it wouldn’t work for general-interest news.

Then the print cliff accelerated faster than anyone anticipated. Between 2008 and 2012, newspaper print ad revenue dropped by roughly half. Digital advertising was growing, but it was growing for Google and Facebook, not publishers. The CPM math was unforgiving: a publisher might earn $25–$40 per thousand impressions in print and $1–$3 online. No volume of page views could close that gap.

I watched this from the inside. When I was publisher of InformationWeek, we had 400,000 controlled-circulation subscribers. These readers received the magazine free because they had the right title, the right company size, the right buying authority. That verified, qualified readership built a rate base: the audited circulation number that determined what we could charge advertisers. The currency from the readers was their time, engagement, and loyalty, which was the entire benefit to advertisers. At its peak, InformationWeek generated over $175 million in annual advertising revenue on that model. The paywall era inverted that logic entirely. It assumed the reader's currency should be cash. For most of publishing history, their attention was worth far more.

The New York Times launched its metered paywall in March 2011, and the industry held its breath. The conventional wisdom was that readers would simply navigate around it, that free content was too available online, that the paywall was a noble but doomed gesture. Instead, the Times started signing up digital subscribers. Not immediately transformative numbers, but real ones. Proof of concept.

Between 2017 and 2020, the number of news outlets implementing paywalls nearly doubled each year. The trade press treated it as a rescue mission.

By the mid-2010s, the paywall had become the dominant narrative of publisher survival. Industry conferences devoted entire tracks to paywall strategy. Consulting firms built practices around it. Publishers who hadn’t implemented one felt pressure to explain why. The enthusiasm was genuine, and in hindsight, genuinely disproportionate to the results most publishers would actually achieve.

Act II: The Scorecard (2018–2023)

Who Won, Who Lost, and Why

Let me give you the honest operator’s scorecard, because the industry press tends to celebrate the wins and bury the losses.

The New York Times won. Decisively. By February 2022, three years ahead of schedule, it hit 10 million digital subscribers and reset its target to 15 million by 2027. But the Times didn’t win because it built a better paywall. It won because it built a better product. Wordle. The Athletic. NYT Cooking. NYT Games. The paywall became a bundle gateway, not a content gate. Subscribers were paying for a suite of daily habits, many of which had nothing to do with news.

The Wall Street Journal won. For the same structural reason I described earlier; it had always been a professional subscription product, and first-party subscriber data made its advertising inventory genuinely more valuable to marketers. WSJ reportedly sells 90% of its inventory direct, and advertisers using its first-party data renew at significantly higher rates. The paywall isn’t a barrier; it’s a data-collection mechanism that funds a premium ad business.

The Financial Times won. Same playbook: a professional audience with employer-funded subscriptions and irreplaceable market-specific content.

Everyone else? Mixed at best. The economics became clear pretty quickly: paywalls work when they protect something a specific reader cannot get anywhere else. General-interest news is available everywhere. Local news is undervalued by local readers until it disappears. Niche B2B content; if it’s genuinely proprietary, or delivered with enough utility that it functions as a workflow tool rather than a reading experience, can command real subscription prices. Bloomberg didn’t build a $6 billion terminal business by having better journalism. It built it by embedding that journalism into a platform traders couldn’t do their jobs without. Some B2B publishers followed a version of that playbook: combining content with data, analytics, or workflow integration until the product became operationally indispensable to its audience. When a reader can’t do their job without you, price sensitivity largely disappears. But that bar is high, and most publisher content — if we’re being honest — never cleared it.

Most households will pay for one or two news subscriptions. That budget overwhelmingly goes to the same three or four scaled national brands. Everyone else is fighting over the remainder.

The Sun tried a hard paywall and abandoned it. The Toronto Star did the same. Countless regional newspapers experimented, retreated, and experimented again. Local news operators found that the readers most willing to pay were often those most engaged. The very people who would have supported the paper anyway. The marginal subscriber who needed to be convinced often just didn’t show up.

Meanwhile, a more structural problem was hardening. Publishers had signed up most of the readers who were ever going to subscribe. Growth was slowing not because the paywall model was wrong, but because the addressable market was smaller than assumed. As one Reuters Institute report bluntly noted, publishers had already captured many of those prepared to pay, and in a tight economic climate it had been hard to persuade others to do the same. Subscription growth was plateauing industry-wide by 2022–2023.

Act III: The Innovation Layer (2022–2025)

Smarter Walls, Shrinking Returns

To their credit, publishers didn’t give up. They got more sophisticated. The blunt instrument of the early paywall — a fixed meter, same for every reader — gave way to dynamic, AI-driven systems that personalize the conversion moment. Instead of offering every reader the same number of free articles, smart meters identify which readers are close to converting and apply pressure accordingly; they identify casual browsers and leave them in the open funnel to generate ad impressions.

The results have been real. Business Insider reported a 75% increase in subscriptions after implementing a dynamic paywall system. The Philadelphia Inquirer achieved a 35% lift in subscriber growth using a similar approach. By late 2024, 38% of news publishers were already using or planning to transition to dynamic paywall models. The internal debate between the subscriber revenue team (lock it down) and the advertising team (keep it open) had found a technology-mediated resolution: optimize for both simultaneously using behavioral data.

This was genuinely encouraging progress. Publishers were finally applying the kind of data discipline to subscription conversion that e-commerce had been using for years. It felt like the industry was growing up.

And then the floor started to move.

Subscribe now

Act IV: The Existential Threat (2025 and Now)

The Wall That AI Walks Right Through

The paywall was designed to stop humans. It was never designed to stop agents.

In October 2025, researchers at the Columbia Journalism Review documented something that should have sent shockwaves through every publisher’s boardroom: AI browsers like OpenAI’s Atlas and Perplexity’s Comet were able to retrieve and summarize a 9,000-word subscriber-exclusive article from MIT Technology Review without triggering any of the barriers designed to block automated access. These systems aren’t scrapers in the traditional sense. They behave like humans. They load the full page, render JavaScript, pass bot-detection systems, because to the website’s infrastructure, they are indistinguishable from a person using Chrome.

The scale is staggering. AI scraping attacks on streaming and media properties jumped 56% year over year. From Q2 through Q4 2025, the rate of AI content scraping grew at an average of 24.4% per quarter. In March 2025 alone, 26 million scraping attempts ignored standard robots.txt directives. The traditional defenses weren’t built for this.

The paywall was designed to stop humans. It was never designed to stop agents. And the agents have arrived.

But the scraping problem, as serious as it is, may actually be the second-order threat. The first-order threat is more subtle and more devastating: if AI systems can synthesize answers from publisher content without sending users to publisher pages, readers never hit the paywall at all. They never become subscribers. The loss is invisible.

Zero-click search rates: queries that get answered without any click to a publisher’s page, rose from 56% in May 2024 to 69% by May 2025. DMG Media reported an 89% drop in click-through rates in September 2025, attributing it directly to AI Overviews in Google search results. The 500 most-visited publishers saw an average traffic decline of 27% year-over-year. Industry analysis estimates AI-powered search summaries reduce publisher traffic by 20% to 60% on average, with niche publications experiencing losses approaching 90%. In an industry where reach equals revenue, this is devastating.

The paywall assumes a reader arrives at your door. Increasingly, they don’t. They get what they needed from the AI and never make the trip.

The Framework: What Actually Works Now

Post-Paywall Revenue Architecture

The lesson isn’t that paywalls failed. It’s that paywalls were always a distribution fix applied to a value problem. Publishers put a gate on content that, in many cases, hadn’t earned the gate. The ones who built sustainable subscription businesses did so because they built something genuinely irreplaceable, then used the paywall to capture value from it.

That logic doesn’t change in the AI era. It intensifies. Here’s what the evidence suggests actually works:

• Proprietary intelligence over commodity content. If an AI can synthesize a reasonable version of your article from publicly available information, your content has a commodity problem, not a paywall problem. The publishers who will survive the AI transition are those with original reporting, exclusive access, proprietary data, or a unique analytical perspective that cannot be reconstructed from digital breadcrumbs. This is what the WSJ has. It’s what the FT has. It’s what most publishers don’t.

• Memberships over subscriptions. The semantics matter. A subscription is a transaction: pay for access. A membership is a relationship: belong to something. The publishers building durable reader revenue are increasingly framing their asks around community, identity, and shared purpose. This is not AI-proof, but it is significantly more human. People don’t cancel memberships the way they cancel subscriptions.

• Experiences over content. Live events, conferences, workshops, dinners — anything that requires human presence cannot be scraped, summarized, or synthesized by an AI agent. This is where B2B media has always had an advantage: the event, the roundtable, the peer-to-peer connection, is genuinely irreplaceable. Publishers who haven’t built an events business need to start building one.

• First-party data as a B2B product. The WSJ model: using subscriber data to make advertising dramatically more valuable, is replicable in principle by any publisher with a real subscriber base and the discipline to develop it. And for B2B publishers who took the utility path, embedding content into workflows, combining editorial with data and analytics, the first-party data advantage is even greater. A reader who uses your product to do their job generates behavioral signals that dwarf anything a casual visitor produces. The challenge is that most publishers let this data sit underutilized, treated as a byproduct rather than a product. In an environment where third-party cookies are dead and AI is eroding traffic-based ad models, the publisher who owns rich, consented, first-party audience data — especially data generated through active, functional use — has a genuine and defensible asset.

• AI licensing deals. The New York Times is suing OpenAI; a case that has escalated into one of the most consequential AI copyright battles in history, now centered on evidence that AI models memorize and reproduce copyrighted content verbatim. The FT signed a licensing deal with OpenAI. Both strategies may prove correct for their respective situations. But the larger point stands: the question every publisher should be asking is not whether to resist AI, but how to monetize their content's role in the AI ecosystem.

The Song Ends the Same Way

The Oasis lyric has one more line worth sitting with: “After all, you’re my wonderwall.” There’s something heartbreaking in it. The conviction that this thing, whatever it is, is the answer. Publishers felt that about paywalls. The conviction was sincere. The timing was real. The economics were genuinely dire enough to justify almost any hopeful narrative.

But a paywall is not a content strategy. It is not a value proposition. It is not a relationship with a reader. It is a gate. And a gate only has value if what’s behind it is worth the price of entry, and if readers can’t get what’s behind it somewhere else for free.

For fifteen years, the publishing industry argued about the gate. The conversation we needed to have was always about what was behind it.

The AI era is forcing that conversation at last. Not because publishers wanted to have it, but because the old evasions have run out. The scrapers are inside the walls. The readers never arrive. The CPMs are gone.

What’s left is the only thing that was ever going to matter: content so good, so specific, so irreplaceable, that a reader would genuinely feel its absence.

Build that. Then put a wall around it.

Subscribe now

The views expressed in Uphoff on Media are entirely my own. They don’t represent the opinions of any company I’ve led, any board I’ve sat on, or any investor who’s had the pleasure of debating strategy with me over the years. If something I write here sounds brilliant, I’ll take full credit. If it turns out to be wrong, I was clearly misquoted by myself.

The most important thing Sequoia Capital published in the last year wasn’t a market map or a fund announcement. It was two sentences:

“The cloud transition was software-as-a-service. The AI transition is service-as-software.”

Ten words. The entire business model of enterprise software: two decades of per-seat licensing, recurring revenue multiples, and stack complexity, reframed in a single breath.

I’ve been running B2B technology and media businesses for 35 years. I’ve lived through the transition from print to digital. From on-premise software to SaaS. From search to social to intent data. Each shift looked, at the time, like a feature upgrade. Each one turned out to be a structural reorganization.

This one is structural. And it’s moving faster than any transition I’ve seen.

The $1/$6 Ratio That Explains Everything

Here’s the number that reframes Sequoia’s thesis from interesting to urgent.

For every $1 a company spends on software, it spends $6 on services.

That ratio — buried in Sequoia’s “Services: The New Software” analysis — is the entire argument. The SaaS era attacked the software market. A $350 billion opportunity. Significant. Transformative. And now largely captured by the incumbents: Salesforce, Microsoft, Oracle, and their horizontal peers.

Agentic AI is attacking the services market. Measured in trillions. And it’s doing it by turning software into a workforce. One that does the work, not just enables it.

This is not an incremental shift. It is a category expansion that dwarfs the cloud transition by an order of magnitude.

What “Service as Software” Actually Means

Let’s be precise, because this concept gets muddy fast.

Software-as-a-Service was about delivery. Software moved from a box you bought to a cloud subscription you rented. The product was still a tool. You still needed people to use it.

Service as Software is about execution. The software doesn’t give your team a better tool. It becomes the team. It does the job. You pay for the outcome, not the access.

Sequoia’s example is Sierra, an AI-based customer support platform. Companies don’t buy Sierra seats. They pay per resolved customer issue. No resolution, no charge. The job-to-be-done gets done, or it doesn’t, and nobody pays. There is no seat. There is no license. There is a result.

That model — pay for work delivered, not software deployed — is the commercial architecture of the next era of enterprise technology.

And it’s already breaking the current one.

The SaaSpocalypse Is Not a Prediction. It Already Happened.

In early 2026, the iShares Expanded Tech-Software ETF fell more than 21% year-to-date. An estimated $2 trillion in market capitalization was erased from B2B software companies in a matter of weeks. Atlassian dropped 36% in a single month. Analysts gave it a name: the SaaSpocalypse.

The market was repricing every software company built on per-seat licensing, because AI agents don’t pay per seat — and neither do the human employees whose jobs they’re replacing. Data emerging from enterprise CIOs showed that for every autonomous AI agent deployed, companies were reducing human software seat requirements at roughly a 1:5 ratio.

The incumbents didn’t argue. They adapted. Salesforce and ServiceNow pivoted to outcome-based pricing, charging for tasks completed rather than seats occupied. That pivot was an acknowledgment, not a strategy. It was recognition that the per-seat model had reached its structural limit.

BCG puts it plainly: 40% of buyers now cite seat reduction as their primary lever to decrease software spending. That’s not a trend. That’s a buyer revolt.

Subscribe now

The CMO Who Signed Up to Do Marketing

Let me bring this down to ground level. Because the abstract inevitably becomes personal.

The average B2B enterprise now runs more than 300 SaaS applications. According to Chief Martec's 2025 landscape report, the martech landscape contains over 15,000 tools. And 32% of organizations report not fully using the capabilities of the stack they’ve already built, up from 28% the year before.

The average CMO did not sign up to be a shadow CTO.

They signed up to do marketing. To build brand. To drive pipeline. To win markets. Instead, they’re spending a material portion of their time managing vendor relationships, integration failures, data reconciliation, and a technology stack that requires its own internal expertise to operate.

This is the hidden tax of horizontal SaaS. The software works. Technically. But operating it at scale, tuned for your specific market and workflow, requires vertical expertise that the horizontal platform was never designed to provide. So you hire it, build it, or hire a partner to build it. And you pay for it twice: once in the license, again in the operating overhead.

Service as Software is the exit ramp from that trap.

When the unit of value is an outcome — leads generated, pipeline influenced, content produced at scale, demand signals identified — the complexity of the underlying technology becomes the vendor’s problem, not yours. The CMO’s job returns to marketing. The vendor’s job becomes delivering the result they promised.

This is not a feature upgrade. It is a reallocation of responsibility.

What This Means for Buyers and Buying Groups

Here’s where it gets structurally interesting, and where most analysis stops short.

The shift to outcome-based procurement doesn’t simplify the buying process. It complicates it in new ways, with new stakeholders.

Today’s B2B buying group already involves an average of 10 people, spanning IT, operations, finance, and end users. 79% of significant purchases now require CFO approval. Procurement professionals — historically a late-stage validator — are now identified as primary decision-makers in 53% of buying cycles, involved from the earliest stages.

When you shift from buying software to buying an outcome, procurement doesn’t get less important. It gets more important. Because now they’re not evaluating features. They’re evaluating a performance contract.

The questions change. “Does it have the functionality we need?” becomes “Can you prove the outcome you’re promising?” “What’s the per-seat cost?” becomes “What are the SLAs and what happens when you miss them?” “Does it integrate with our stack?” becomes “How do you measure success and how do we audit it?”

This raises the analytical bar for every vendor in the room. And it raises it specifically for B2B marketing technology vendors, because marketing outcomes have historically been the hardest to attribute and the easiest to obscure.

Service as Software forces accountability. That’s good for buyers. It’s clarifying for vendors who can deliver. It’s existential for vendors who can’t.

The Buying System Is Now Machine + Human

There’s a dimension to this shift that almost no one in B2B is taking seriously yet, and it changes the implications of everything above.

The buyer isn’t becoming a machine. The buying system is becoming machine-human integrated.

Gartner projects that AI agents will be involved in the vast majority of B2B purchases within three years, channeling more than $15 trillion in spending through automated evaluation. Forrester found that 94% of B2B buyers already use AI in the purchasing process, and that twice as many buyers now name generative AI as a more meaningful information source than vendor websites or sales teams.

But here’s the critical nuance: buyers aren’t handing off to agents and waiting for a shortlist. They’re working with agents iteratively; prompting, refining, validating, overriding. The agent surfaces structured signals. The human interprets, challenges, and decides. The agent adjusts. The cycle repeats.

By the time your Service as Software offering reaches a human conversation, it has already been filtered, scored, and contextualized by a machine that the human trusts enough to be working with in the first place.

You are not marketing to a human audience anymore. You are marketing to a system. And that system has two components that must be satisfied simultaneously, not sequentially.

The implications for how you go to market are significant.

The machine component of the buying system doesn’t respond to corporate narrative. It evaluates structured, verifiable, consistent proof. “A case study stating ‘improved productivity’ is invisible to agents,” Forrester noted in its 2026 State of Business Buying report. “Reduced processing time from 14 days to 3 days is specific enough to extract and compare.”

The human component, working alongside an increasingly capable agent, arrives at vendor conversations better prepared and less tolerant of information they’ve already sourced. They don’t need you to explain the category. They need you to demonstrate judgment and point of view that the agent couldn’t surface on its own.

If your pricing is opaque, your outcomes aren’t expressed in quantifiable terms, or your proof infrastructure is thin, you may lose the deal before anyone picked up the phone. And you’ll never know it happened.

What This Does to Partner Marketing

Most channel marketing strategies were built for a licensing economy.

The reseller partner who moved seats earned margin on the transaction. The systems integrator who deployed the platform earned margin on the implementation. The referral partner who brought the deal earned a percentage of ACV. Every incentive in the traditional partner model pointed at the software transaction.

Service as Software decouples value from the transaction. If the vendor is now responsible for the outcome, the traditional partner’s role collapses. Moving a license to a customer who then fails to get results isn’t a win. It’s a liability.

The partner that wins in a Service as Software economy is a different entity entirely: one that configures, manages, and guarantees the outcome on behalf of the buyer. A delivery partner, not a distribution partner. Accountable for outcomes, not for moving licenses. The ISV relationship becomes a managed service overlay on top of agentic capability, not a distribution arrangement.

For B2B marketing leaders managing partner programs: this is not an edge case. It is the direction of travel. Your partner tiering, your co-sell incentives, your MDF programs, all were designed for a software transaction model. Most will need to be rebuilt for an outcome delivery model.

The Transition Is Messier Than the Thesis

I want to be clear about something, because this is an operator’s publication and operators live in reality.

The SaaSpocalypse was the market pricing the transition’s messiness in real time. Two trillion dollars of repricing in a matter of weeks is the capital markets saying: we don’t know which SaaS companies survive this shift, so we’re discounting all of them until they prove it. That’s not irrational. It’s pattern recognition from investors who’ve seen this movie before.

But the actual operating transition is slower and more hybrid than the market event implies.

According to a 2025 SaaS Pricing Benchmark Study, only 9% of companies have fully implemented outcome-based pricing. 47% are actively exploring or piloting it. The remaining 44% are still running the old model, and will be for some time.

BCG expects that most mature vendors will combine models: some subscription baseline, some agent-based pricing, some outcome-based performance fees. Pricing structures will be more complex before they’re simpler. Sales motions will be more consultative before they’re more automated. Buyers will need new RFP frameworks, new measurement practices, and new governance structures before they can fully operationalize outcome purchasing at scale.

And Gartner has flagged that at least 30% of GenAI projects will be abandoned due to unclear business value. The hype cycle is real. Some of what’s being sold as “Service as Software” today is simply repackaged SaaS with a new name on it.

The structural direction is clear. The pace is not. Know which one you’re operating in.

What Operators Need to Do Now

For B2B marketing leaders:

Buy outcomes, not access. Start by auditing your existing stack against a single question: is this vendor making my team faster, or making my team unnecessary? Those are different things, and only one of them is the future. The answer tells you which vendors survive your next renewal cycle, and which conversations you should already be having.

Develop RFP frameworks that demand outcome commitments, not feature lists. Build the internal capability to measure vendor performance against those commitments. If you can’t measure it, you can’t hold anyone accountable for it.

And structure your proof assets for the Machine + Human buying system, not just for a human audience. Quantifiable, verifiable, structured outcomes aren’t just good storytelling. They’re the minimum table stakes for being found and evaluated by the system that precedes the human conversation.

For B2B technology vendors:

The seat count is no longer a growth metric. Outcomes delivered, outcomes measurable, outcomes attributable, those are your new unit economics. If you cannot articulate what success looks like in quantifiable terms and stand behind it contractually, you are not positioned for this transition.

If your go-to-market motion is still feature-benefit selling, you are selling the wrong thing to a buyer whose purchasing system has already moved on.

For B2B media and marketing service providers:

This is your moment. If you can execute. The demand for outcome-based marketing services has never been higher. Content syndication, demand generation, account-based programs: all become far more valuable when they’re measured against pipeline outcomes and delivered at scale with agentic efficiency.

The providers who can price to outcomes and prove them will displace the vendors still selling impressions, leads, and seats.

The Bigger Picture

The SaaS era gave buyers better tools. The Service as Software era gives buyers better results, or it doesn’t get paid.

That accountability shift is a reorganization of incentives across the entire B2B technology ecosystem. When vendors succeed only when customers succeed, the dynamics of product development, customer success, go-to-market, and partner strategy all change. The consultative relationship that enterprise vendors aspired to becomes a contractual requirement.

For the CMO who has spent five years managing a 25-platform stack, wondering when they stopped doing marketing and started running a technology operation, that reckoning is coming. And it’s going to feel like relief.

The next generation of B2B marketing leaders won’t manage a martech stack. They’ll manage a portfolio of outcomes. And they’ll have a lot more time to actually do marketing.

The transition is underway. The direction is clear. The only question is how fast you adapt, and whether you’re buying the real thing or the repackaged version of the old model with a new name on it.

Subscribe now

The views expressed in Uphoff on Media are entirely my own. They don’t represent the opinions of any company I’ve led, any board I’ve sat on, or any investor who’s had the pleasure of debating strategy with me over the years. If something I write here sounds brilliant, I’ll take full credit. If it turns out to be wrong, I was clearly misquoted by myself.

Gartner’s latest research: buyers spend just 17% of their total purchasing time in direct contact with vendors. 83% define their requirements before talking to sales. 81% already have a preferred vendor when they finally engage a rep. The buying journey isn’t a funnel anymore. It’s a verdict with an appeals process attached.

The conventional worry about all of this is disintermediation. If buyers reach shortlist without talking to sales, what exactly is sales for?

That’s the wrong question.

The right question is this: when a buyer arrives at 70 or 80 percent of the journey, who is carrying the risk?

It’s not the rep.

The Asymmetry Nobody Talks About

Consider what a CFO, CTO, or Chief Procurement Officer has personally staked by the time they engage a vendor. They’ve built an internal case. They’ve put their credibility on the line with their board, their CEO, their peers. They’ve made a professional judgment about which vendors belong on the shortlist and why. 86% of B2B purchases stall during the buying process, and when they do, someone inside the buying organization owns that failure. If the decision goes wrong — if the implementation struggles, the promised outcomes don’t materialize, the vendor relationship deteriorates — they are exposed. Careers have ended over enterprise software decisions gone bad.

The sales rep carries none of that exposure. If the deal closes and the customer struggles, the rep has moved on to the next quarter. The asymmetry is structural. And it is rarely acknowledged.

This isn’t an indictment of enterprise sales. It’s a design flaw that the industry has been slow to recognize. In the old model, the one enterprise sales was built for, the human relationship formed early, developed over time, and quietly absorbed much of this risk. The buyer didn’t just trust the vendor. They trusted a person. “I know this rep. They’ll go to bat for us. They know what we need.” That relationship was a rational risk transfer mechanism. The rep’s career incentive, their account ownership, their reputation, all of it created an informal backstop for the buyer’s judgment.

Strip away that early relationship, and the risk doesn’t disappear. It stays with the buyer. And now they’re carrying it alone, built on a thesis assembled from digital signals, with no human ally who helped them form it.

When the Story Doesn’t Match

Here’s where it gets more complicated, and where marketing enters the picture in a way most companies haven’t fully reckoned with.

During that 70 to 80 percent journey, the buying committee isn’t operating in an information vacuum. 70% of buyers say they prefer to learn about a product or service through content — blogs, videos, case studies — rather than traditional sales outreach. They are consuming executive interviews, keynote videos, thought leadership posts, product narratives, analyst briefings. 47% of buyers view three to five pieces of a company’s content before talking to a sales representative.

By the time the rep gets the meeting, the buying committee has already formed a relationship. Not with a person, but with a story.

What happens when the rep’s narrative doesn’t match that story?

Not a jarring, obvious contradiction. Something subtler. A different emphasis. A different vocabulary. The marketing narrative positioned the platform around operational transformation; the rep leads with cost reduction. The exec videos talked about partnership and long-term outcomes; the rep is focused on closing the quarter. The content established one set of expectations about how this company engages; the rep’s behavior signals something different.

The buying committee doesn’t diagnose this as a sales and marketing alignment problem. They experience it as unease. Something feels off. In a low-stakes purchase, that feeling gets rationalized away. In a high-stakes enterprise decision — where careers are on the line and 79% of enterprise deals require final CFO approval — it doesn’t. It becomes a quiet reason to slow the process, reopen the evaluation, or move toward a competitor who feels more coherent.

The deal doesn’t die loudly. It just stops moving.

Ground Truth

The most sophisticated enterprise marketing leaders are starting to name this problem explicitly. One framework, shared with me by a senior marketing executive at a major enterprise software company, adds a third dimension to the traditional brand and market truth model. Brand truth is what the company believes about itself. Market truth is what analysts, press, and the broader market reflect back. Ground truth is what buyers actually experience and conclude through their own research journey and customer experience.

The insight is that misalignment between these three isn’t just a messaging problem. It’s a trust problem. And in a digital-first buying journey, trust failures are harder to detect and harder to recover from, because there’s no human relationship in place to spot them and absorb the friction.

This is the same dynamic explored in “B2B Branding Has a New Audience” but where that post examines it from the brand and marketing side, the sales consequences are what make it existential. When agentic AI exposes messaging drift across every digital surface, the rep walking into that 80 percent meeting doesn't just carry narrative inconsistency into the room. They’re closing a gap that an algorithm has already scored against them.

For enterprise marketing, the implication is structural. Every content asset, every executive communication, every campaign narrative needs to be built with the rep’s eventual conversation in mind. Not as a constraint on creativity, but as a discipline of coherence. The rep is the last mile of the content journey. 82% of B2B buyers say thought leadership from individuals, not just corporate messaging, influences their purchasing decisions. That makes every exec video, every LinkedIn post, every keynote a load-bearing element of the sales motion. If the last mile contradicts the miles that came before it, the buyer notices. Even if they can’t articulate exactly what they noticed.

And that framework is about to face a stress test nobody designed it for.

Agentic AI Will Make This Worse Before Anyone Notices

Here is the part of this story that isn’t being talked about yet.

Agentic AI is entering the enterprise buying journey, and it is amplifying every tension described above. 90% of procurement executives say they have considered or are actively planning to use AI agents to optimize their procurement operations. Gartner projects that the majority of enterprise software applications will include agentic AI by 2028. Several major procurement platforms; SAP, Coupa, and others, have begun embedding these agents natively. The planning phase is collapsing into deployment. This isn’t a distant future. It’s the next procurement cycle.

What does an agentic AI procurement system actually do? It doesn’t wait for a human to run a search. It autonomously conducts vendor research, synthesizes reviews and analyst reports, scores vendors against predetermined criteria, and surfaces a ranked shortlist. Often before a human buyer has had a single conscious thought about the category. This is the B2A shift — Business to Algorithm — and it’s not emerging. It’s here. Companies that lack clean, structured, algorithm-ready data risk being invisible in digital buying journeys before a human ever enters the picture. Salesforce’s Agentforce and a growing category of purpose-built procurement AI systems are moving this from thought experiment to Q3 budget conversation.

This changes the risk calculus in a way that nobody in enterprise sales has fully processed.

When a human buyer conducts self-directed research, they are forming impressions, making judgment calls, weighing intangibles. There is still a human in the loop who can be reached, influenced, and reassured. When an AI agent conducts the initial research and builds the shortlist, the process is faster, more systematic, and far less transparent. The buying committee receives a ranked list of vendors, and critically, 83% of buyers modify their initial shortlist after conducting further research, with more than a quarter making significant or complete changes. But if the initial list was built by an agent operating on criteria the rep never knew existed, the vendor who scores poorly may never understand why.

The “something feels off” problem gets structurally worse. When a human buyer felt unease, a skilled rep could sometimes read the faint signals, ask the right question, and surface the concern. When an agent scores your company’s narrative consistency across fifty digital touchpoints and weights it against competitors, the friction is invisible. The deal just doesn’t progress. No signal. No conversation. No recoverable moment.

This is why the Ground Truth framework matters more now than it did even two years ago.

A human buyer might feel it. An AI agent will score it. And it will do so at a stage of the journey so early that the enterprise sales rep will never know the conversation was happening.

How Enterprise Sales Actually Needs to Evolve

The rep entering a 70 to 80 percent buyer journey is no longer a pitcher. They are a validator…and more precisely, a risk translator.

The buyer has built a thesis, often over months, increasingly with AI assistance. The rep’s first job is to understand it, confirm what’s right, and carefully address what isn’t. Not to restart the narrative from the company’s preferred talking points. The buyer has invested heavily in their own conclusions. A rep who ignores that investment creates resistance. A rep who honors it earns trust.

Beyond validation, the rep needs to understand what the buyer has personally at stake. Risk qualification matters as much as lead qualification. What has this buyer committed to internally? What would a poor outcome mean for them specifically? That isn’t soft skills territory. It’s the core intelligence that determines how the rep should engage, what assurances matter most, and where the deal is actually fragile.

The rep’s evolved role is risk translation: helping the buyer convert the thesis they’ve built alone into a defensible internal commitment, with a vendor partner who will stand behind it. That’s the trust transfer the old model delivered through relationship. Now it has to be earned in a fraction of the time, at the back end of a journey the rep didn’t take with them.

At the organizational level, the companies that win in a digital-first world will treat sales and marketing alignment not as a coordination exercise but as a buyer experience imperative. The story the company tells; across every format, every channel, every exec communication, needs to be the same story the rep walks in with. Not identical in script, but coherent in thesis, consistent in values, aligned in vocabulary. Because that story is now being processed not just by human buyers making judgment calls, but by AI systems making scores.

The Quiet Deals You’re Losing

The buyer seemed engaged. The evaluation went well. Then it stalled. Then it went quiet.

Often, something felt off. The buyer couldn’t name it. The rep never knew. And increasingly, the agent that built the initial shortlist made a call no human on either side of the table ever saw coming.

In a high-stakes purchase made by humans who are personally exposed, coherence isn’t a nice-to-have. It’s a closing condition. The companies that figure this out (that the rep is the last mile of a journey the buyer took largely alone, that the buyer arrived carrying real personal risk, and that AI agents are now entering that journey at its earliest and most consequential stages) will win deals their competitors don't even know they lost.

Enterprise buyers have changed. The companies that change with them will define the next era of enterprise sales.

Subscribe now

The views expressed in Uphoff on Media are entirely my own. They don’t represent the opinions of any company I’ve led, any board I’ve sat on, or any investor who’s had the pleasure of debating strategy with me over the years. If something I write here sounds brilliant, I’ll take full credit. If it turns out to be wrong, I was clearly misquoted by myself.

Not a few readers. Enough that I noticed a pattern. The request was consistent, and specific. People wanted to listen while driving, traveling, working out. They were already reading the newsletter; they wanted the same content in a format that fit the rest of their day.

You asked. Here it is.

Uphoff on Media is now available as a podcast on Apple and Spotify. Each episode is the audio version of the newsletter, the same analysis, the same operator perspective, the same thesis-first take on the trends that are reshaping B2B media, marketing, and technology.

If you’ve been reading, you already know what you’re getting. This is just a different way to take it with you.

Listen on Apple Podcasts →

Listen on Spotify →

My thanks to the readers who suggested this. If you find it valuable, a rating or review on Apple Podcasts is the single best thing you can do to help a new show find its audience. Thanks.

Let’s get to work

Tony

Subscribe now

B2B brand building is about to be restructured from the ground up.

Not incrementally refined. Not optimized. Structurally inverted.

Here’s the reality most B2B CMOs haven’t fully reckoned with yet: in the Agentic AI era, your brand is no longer being evaluated by a human audience with machine tools. It’s being evaluated by a Machine + Human system. Agents and buyers working together, simultaneously and interdependently, throughout the entire buying journey.

The enterprise buyer isn’t handing off to an AI agent and waiting for a shortlist. They’re working with the agent iteratively, prompting, refining, validating, overriding. The agent surfaces structured signals. The human interprets, challenges, and decides. The agent adjusts. The cycle repeats. By the time your brand reaches a human conversation, it has already been filtered, scored, and contextualized by a machine that the human trusts enough to be working with in the first place.

You are no longer positioning to a human audience. You are positioning to a system, and that system has two components that must be satisfied simultaneously, not sequentially.

Everything about how B2B brands are built, measured, and differentiated follows from that premise.

The Fuel Mix Is Changing

For decades, B2B brand building ran on a relatively stable fuel mix: analyst relationships, trade media coverage, event presence, demand generation programs, and thought leadership content. Each of these was designed for human consumption; a buyer reads, watches, attends, engages, and forms impressions over time.

That mix isn’t going to zero. But the weighting is shifting dramatically, because the audience has changed.

The machine component of the buying system doesn’t respond to corporate narrative. It evaluates structured, verifiable, consistent proof. That means work most marketing organizations have never treated as brand work: schema markup, review platform completeness, certification maintenance, consistent metadata across every surface where an agent might evaluate you, is now core brand-building activity. Not IT work. Not legal work. Brand work.

The human component of the buying system is simultaneously getting more sophisticated and more selective. Buyers working with AI agents arrive at human conversations better prepared, with higher expectations, and less tolerance for information they’ve already sourced themselves. They don’t need you to explain the category. They need you to demonstrate judgment, credibility, and point of view that the agent couldn’t surface on its own.

Building for both components, at once, coherently, is the new mandate.

Trust Becomes the Dominant Variable — But It Now Operates as a System

In every era of market disruption, trust becomes more valuable. When uncertainty is high and the pace of change is relentless, buyers look for the trust signal that most reliably predicts competence and consistency.

This moment is no different. But the agentic era changes how trust is built and evaluated, because trust now has to work across both components of the buying system simultaneously.

Machine trust is built on verifiability. Quantitative ratings, independent certifications, contractual transparency, uptime records, security documentation. Agents weight these signals heavily because they’re legible, consistent, and difficult to fake at scale. You either have them or you don’t. No amount of compelling brand narrative compensates for weak machine trust signals, the agent simply doesn’t register them.

Human trust is built on authority, credibility, and resonance. Does this vendor think clearly? Do they understand my problem better than my own team does? Do I trust their judgment in a world that’s moving too fast to validate everything myself? This is the trust that gets built through thought leadership, practitioner voice, executive presence, and community standing. No amount of machine-legible proof compensates for weak human trust. The buyer ratifies the agent’s shortlist, not the other way around.

There’s a reason B2B brand advertising has always underperformed. Most marketers treat enterprise purchasing as a rational process: feature sets, integration specs, pricing tiers, ROI models. They optimize for the quantitative case and underinvest in the emotional one. But B2B purchase decisions have never been purely rational. Risk is emotional. Trust is emotional. The unspoken question in every enterprise evaluation — if this goes wrong, can I defend this choice? — is entirely emotional. Nobody ever got fired for buying IBM wasn’t a product claim. It was an insight into how fear shapes procurement. In the agentic era, this dynamic doesn’t diminish. It concentrates. The machine handles the rational layer: structured proof, verified signals, consistent metadata. What remains for the human is judgment, confidence, and the emotional calculus of professional risk.

Brand has always lived there. Now it’s the only place left that machines can’t reach.

But here’s the critical point the sequential framing misses: these two forms of trust don’t operate in separate phases. They interact. A human buyer working with an agent is weighing both simultaneously, asking the agent to pull structured data while applying their own judgment about which vendors feel credible.

A message that works perfectly for an agent but lands wrong with the human using that agent is still a failure. The signals have to be coherent within the system, not just adequate on each dimension independently.

The brands that win are the ones where machine trust and human trust are strong, aligned, and mutually reinforcing. That alignment is the new core challenge for B2B marketers.

Agentic AI Will Expose Every Inconsistency You’ve Been Living With

Here’s the near-term operational reality most B2B marketing leaders haven’t confronted yet: your brand is probably already inconsistent across the surfaces that matter. Not because of bad strategy. Because of organizational entropy.

Your website says one thing. Your G2 profile says something slightly different. Your LinkedIn company page hasn’t been updated since the last rebrand. Your partner listings describe a product you pivoted away from eighteen months ago. Your pricing page is deliberately vague while your sales deck quotes specific numbers. Your analyst profile reflects a positioning you outgrew two years ago.

In the human-only era, buyers navigated this inconsistency intuitively. They triangulated, asked questions, and gave you the benefit of the doubt. Inconsistency was sloppy but survivable.

In the Machine + Human era, it’s a trust penalty. A compounding one.

When the machine component of the buying system evaluates your brand across multiple surfaces — and it will — conflicting signals don’t average out. They create doubt. An agent trained to identify reliable vendors treats inconsistency as a risk signal, not a nuance to be interpreted generously. Misaligned messaging, contradictory value propositions, opaque commercial terms, these don’t just confuse the machine. They move you down the shortlist, or off it entirely. And a human buyer who sees those inconsistencies flagged, or encounters them directly, loses confidence in ways that are very hard to recover from mid-journey.

The immediate tactical implication: the CMO who conducts a rigorous brand consistency audit across every machine-legible surface right now; website, review platforms, partner listings, analyst profiles, press coverage, social presence, commercial documentation, has a near-term competitive advantage that costs almost nothing to capture. This isn’t glamorous brand strategy work. It’s the operational prerequisite for everything else in this post.

Do it before your competitors do.

Subscribe now

Storytelling Goes Up in Value, Down in Volume

Here’s the counterintuitive move. In a world where the machine component of the buying system handles discovery and structured evaluation, narrative and storytelling don’t become less important. They become more important, and more precisely targeted.

When AI commoditizes the discovery and evaluation layer, what remains as true differentiation in the human conversation? Point of view. Intellectual honesty. The ability to frame a problem so precisely that a buyer feels understood in ways no structured data source could.

Narrative becomes the currency of the final mile. The element of your brand that the machine can surface evidence of, but can’t replicate or replace.

The catch: the human component of the buying system is smaller and more consequential than the broad human audience of the previous era. Fewer people are in the room. So storytelling’s value per impression goes up while its distribution volume goes down. The implication for content strategy is significant: invest in fewer, deeper, more operator-credible pieces rather than volume content designed to feed an automated pipeline. In a Machine + Human buying system, depth and credibility outperform volume every time.

Where to Invest, Where to Pull Back: The Channel Mix Is Being Reordered

Some of this is already in motion. The agentic era accelerates and amplifies each of these shifts, because every medium now has to be evaluated through the lens of what it contributes to the Machine + Human buying system, not just to human awareness.

↑Experiential events: Up. When the machine component handles discovery and initial evaluation, live human presence becomes the scarcest and highest-fidelity signal in the market. In-person connection is where machine-shortlisted vendors become trusted partners. Where the human component of the buying system forms the judgments that structured data can’t. The trade show model continues to struggle. The curated dinner, the practitioner roundtable, the executive cohort — these go up. Not because they’re nostalgic, but because they operate precisely in the space the machine can’t reach.

↔Demand generation: Restructured, not eliminated: The form fill, the gated asset, the email drip sequence, these specific mechanics are structurally misaligned with a Machine + Human buying journey. But the underlying function of demand generation doesn’t go away. What changes is the model.

B2B marketers have hit a genuine tipping point with technology complexity. The martech stack — built layer by layer over the past fifteen years — was designed for a world of human buyers moving through discrete, trackable stages. That world is ending. What replaces it isn’t more technology. It’s a fundamentally different service model.

The demand generation providers who survive and lead in the agentic era will deliver something closer to Demand-as-a-Service: multi-channel engagement, pipeline impact, and analytics integrated into a single coordinated offering rather than a collection of separate tools and tactics stitched together by an overtaxed marketing team. The separation of “content agency” from “media buyer” from “analytics platform” from “SDR team” is a legacy of a simpler era. Agentic AI collapses those distinctions, or more precisely, it makes the cost of maintaining them prohibitive.

Successful demand generation in the agentic era requires coordinated engagement across both the machine and human components of the buying journey, sequenced intelligently rather than executed as parallel, disconnected programs. An agent shortlists you. A human validates. A practitioner-influencer confirms. A curated event closes. That’s a coordinated buyer journey, and it requires a coordinated provider, not five vendors trying to hand off to each other.

↑Thought leadership: Up, but restructured: The white paper with three industry endorsements and a PDF download is finished. What replaces it: structured POV content that’s simultaneously machine-indexable and human-resonant. Short, declarative, operator-credible, and tied to a genuine point of view about the market. The executive voice matters more now, not less, because it’s one of the few signals that works meaningfully for both components of the buying system at once. But it has to be real, not ghostwritten corporate prose.

↑B2B influencers: Significant rise: Not celebrity influencers. Practitioners-as-influencers. People who have actually done the job, carry functional credibility with your buyer, and speak in a register that earns trust precisely because it doesn’t sound like marketing. This signal works across both components of the buying system: machine-indexable as third-party validation, and human-resonant as peer credibility. That dual utility makes it disproportionately valuable. Expect it to become a meaningful budget line.

Academia Is Waning, And That Trend Accelerates

For decades, academic credibility was a core pillar in B2B tech markets. Research papers, university partnerships, PhD-certified methodologies, these were proxies for rigor and third-party validation in a world where buyers had limited tools for independent verification.

That proxy is breaking down, and the Machine + Human buying system accelerates the breakdown from both directions.

The machine component can now source, cross-reference, and evaluate far more diverse evidence than any buyer could manually. Academic affiliation is just one signal among many, and not a particularly fast-moving one. The cycle time problem alone is disqualifying: academic research takes years; B2B markets now move in quarters. The human component, working alongside increasingly capable agents, values current, contextual, operator-grounded analysis over credentialed but dated research.

What fills the gap: analyst-practitioners, operator thought leaders, and community-validated research. The rise of independent expert voices — people with track records, real stakes, and current market engagement — isn’t a trend. It’s a structural shift in how both components of the buying system evaluate what to believe and trust. The machine surfaces them. The human recognizes them. Together, they move the needle in ways institutional research no longer can.

New Tools Are Coming, And the Category Doesn’t Exist Yet

If brand trust now has to operate coherently across a Machine + Human buying system, the diagnostic tools have to change too.

The current marketing analytics stack was built to measure human behavior. Impressions, clicks, downloads, MQLs, pipeline attribution. Those tools don’t answer the questions that matter now:

• When an AI agent queries for solutions in my category, does my brand appear? With what frequency, in what context, with what sentiment?

• How complete, consistent, and machine-readable is my proof infrastructure across every surface the buying system evaluates?

• Does my brand tell the same story across website, G2, LinkedIn, partner descriptions, analyst profiles, and press coverage, or are there inconsistencies creating trust penalties in the machine layer?

• What percentage of credible practitioner content in my category validates or references my brand, and is that content machine-indexable as well as human-credible?

• And eventually, the killer metric: what percentage of agentic buying journeys in my category end with my brand on the shortlist?

Nobody has built this cleanly yet.

Early signals are emerging, companies like Profound are taking initial runs at AI visibility measurement, and some SEO intelligence firms are beginning to pivot from search visibility to agent visibility. The tool that measures machine trust, human trust, and the gap between them doesn't exist yet.

That's not a problem. That's a market opportunity.

The firm that builds a credible Agentic Brand Trust Score; something a CMO can track quarterly, benchmark against competitors, and act on across both dimensions, will define a significant new category of B2B marketing infrastructure. It’s the next essential instrument in the CMO’s toolkit, and it will emerge from AI-native analytics startups rather than from the established marketing technology vendors who built their platforms for a different world.

New Firms Will Win This Market. Here's What They Look Like.

I’ve written about where traditional ad agencies are headed. The creative-driven, brand-narrative firm built for human media consumption is structurally misaligned with what B2B marketers now need, because it was built for one component of a system that now has two.

Five capabilities the market will pay for that barely exist today:

Structured narrative architects build brand signals that work simultaneously for machine legibility and human resonance. This sits at the intersection of content strategy, data architecture, and brand thinking, and it doesn’t really exist inside any current agency or consulting model.

Agentic content strategists understand how AI agents discover, evaluate, and weight vendors, and design content specifically to perform well within that process. This isn’t SEO. It’s something more fundamental: optimizing for how the machine component of the buying system thinks, while maintaining the human credibility that earns the final decision.

Operator thought leadership producers help executive teams develop and distribute credible, practitioner-voice content at scale without sanitizing it into corporate messaging. The voice that works for both components of the buying system is rare. Producing it consistently is a real capability gap.

Trust infrastructure consultants audit and build the proof-point architecture; reviews, certifications, structured data, third-party validations, pricing transparency, that the machine component of the buying system actually evaluates. This work currently sits in no-man’s-land between marketing, IT, and legal. The firm that owns it will be valuable.

Community architects build owned audiences that aren’t dependent on algorithms, platforms, or machine discovery. In a world where automated shortlisting is the default, a trusted community of engaged practitioners is a durable moat, and a source of human trust signals no machine can manufacture.

None of these exist inside a current agency or consulting firm. That's the point.

The Strategic Imperative

The brands that navigate this transition well will understand something most of their competitors won’t.

This isn’t a channel shift. It’s not a new media mix or a technology upgrade. It’s a structural change in the nature of the audience for B2B brand signals: from a human audience assisted by tools, to an integrated Machine + Human system in which both components must be satisfied coherently and simultaneously.

Building for that reality requires rethinking brand strategy, content strategy, measurement infrastructure, and the partner ecosystem at the same time. That’s hard work. Most organizations will do it slowly, partially, and reactively.

The ones that do it deliberately, that treat the Machine + Human buying system as the design constraint for everything they build, will have a compounding competitive advantage that gets harder to close over time.

The question every B2B CMO needs to answer right now isn’t whether this shift is happening. It’s whether they’re building a brand architecture designed for the buying system that exists today, or still optimizing for the one that existed five years ago.

Subscribe now

The views expressed in Uphoff on Media are entirely my own. They don’t represent the opinions of any company I’ve led, any board I’ve sat on, or any investor who’s had the pleasure of debating strategy with me over the years. If something I write here sounds brilliant, I’ll take full credit. If it turns out to be wrong, I was clearly misquoted by myself.

“Culture eats strategy for breakfast.”

Leaders have invoked it for decades as a reminder that the human architecture of an organization ultimately determines what’s possible, regardless of how elegant the strategy is. Get the culture wrong and the strategy dies on the whiteboard.

But here's what Drucker didn't anticipate: the rise of agentic AI — and what happens when the culture itself is the most disrupted group? When the people who built your culture, who carry it, who embody it every day, are the same people most at risk from the change you’re asking them to embrace?

That’s the leadership challenge of the Agentic AI moment. And it’s unlike anything executives have navigated before.

The challenge of leading organizational culture through AI transformation is now the defining test of B2B executive leadership — and most change management frameworks weren't built for it.

Why Traditional Change Management Fails with AI Disruption

Every major disruption of the past three decades; the internet, mobile, cloud, the 2008 financial crisis, tested organizational cultures. Leaders learned hard lessons about change management, communication, and the importance of bringing people along. Entire frameworks were built around it. Kotter’s 8-Step Model. ADKAR. Prosci. The consulting industry made billions helping companies implement them.

These frameworks share a common assumption: the workforce is the response team. The disruption arrives from outside — a new competitor, a market collapse, a technological shift — and the leader’s job is to rally the organization to meet it. Culture is the obstacle to navigate or the asset to leverage. Either way, the people are with you, not in the path of what's coming.

These frameworks were built for a world where the workforce was the response team. Not the most disrupted constituency.

Agentic AI breaks this assumption entirely.

In many organizations, the largest portion of the workforce, people in operational, administrative, analytical, and content roles, are the ones most directly in the path of AI-driven automation. They're not the response team. They're the most affected party. And you’re asking them to help lead the transformation.

The usual playbook was never designed for this. Which is why so many leaders are finding that their best-intentioned efforts — the town halls, the AI task forces, the “we’re investing in our people” messaging — aren’t moving the needle on culture the way they expected.

They’re not failing because they’re doing it wrong. They’re failing because they’re solving the wrong problem.

The Trust Paradox: Why Employees Don't Believe AI Reassurances

Here is what actually happens to trust during major disruptions, and why most leaders misread it.

When organizations face extraordinary change, employees demand more from their leaders. More transparency, more honesty, more reassurance. Trust becomes the dominant cultural conversation. Leaders respond the way they’ve been trained to: they work to rebuild the employer-employee relationship. More communication. More visibility. More town halls. More open-door moments.

And it doesn’t work. The trust metrics don’t improve. The anxiety doesn’t subside. The cultural resistance hardens.

The reason is this: the trust that’s actually eroding isn’t trust in the company. It’s trust in the future. Specifically, it’s employees’ trust that their skills will remain relevant, that their career trajectories will survive, that the professional identities they’ve built over years or decades will still mean something on the other side of this transition.

That’s not a relationship problem. No amount of authentic communication or leadership visibility repairs it. Because the honest answer to “will I be okay?” is one no leader can give with certainty — and employees know it.

Chart 1: The Trust Paradox

This is the Trust Paradox. Employees are losing trust in the future and expressing it as distrust of their employer. Leaders diagnose a relationship problem and try to fix it with relationship tools. The gap between the intervention and the actual need widens. And employees — who know exactly what they’re afraid of — lose even more confidence when they see leadership dancing around the real question.

I watched this play out with unusual clarity when I was running UBM TechWeb during the 2008 financial crisis. We weren’t managing one disruption. We were managing four simultaneously: the structural decline of print, a digital business still proving its economics, an events portfolio we were actively acquiring and globalizing, and a workforce spread across the US, UK, India, and China. The anxiety inside the organization wasn't primarily about the financial crisis, as severe as that was. It was about the medium. Print journalists and advertising sales professionals who had built twenty-year careers were watching their professional world contract in ways that felt permanent — and they were right. We were bringing in new digital leaders, launching new digital products, making a sharp strategic pivot toward events. No market recovery was going to reverse any of it.

They didn't need reassurance. They needed something more honest, and more useful.

What I learned leading through that period: the leaders who navigate disruption successfully aren’t the ones who rebuild trust most effectively. They’re the ones who stop trying to fix the wrong problem.

First Move: Radical Transparency About AI's Impact

The most counterintuitive thing a leader can do right now is also the most important: say out loud what everyone already knows.

Your employees are reading the same headlines you are. They’re using AI tools at home. They’re having conversations with colleagues, partners, and recruiters that you’re not in the room for. The disruption is not a secret. The anxiety is not hidden. And any attempt to soften, sequence, or carefully manage the narrative reads — accurately — as avoidance.

Transparency in this context isn’t a communication strategy. It’s a leadership stance. It means walking into the room and saying: “here is what we know about what Agentic AI will change in this organization. Here is what we genuinely don't know yet. And here is how we intend to navigate it — together”.

That last word carries weight. Not “here’s what we’re doing to you.” Not “here’s what we’re doing for you.” Here’s what we’re navigating together.

Change at this scale cannot be snuck up on. The disruption is too visible. The anxieties are rising too fast. Leaders who wait for the “right moment” to have the honest conversation are ceding the narrative to rumor, speculation, and fear, all of which fill the vacuum faster than any town hall can address.

The counterintuitive truth: naming the disruption explicitly, including the uncertainty, builds more durable credibility than projecting confidence you don’t fully have. Because when you subsequently make the case for where you’re going and why, people believe it. You haven’t asked them to disbelieve what they can already see.

Name it. Don’t manage it.

Second Move: Build Navigational Confidence, Not Just Trust

Once you’ve named the disruption honestly, the next move is equally important — and equally counterintuitive.

Stop trying to rebuild trust. Start building confidence in the organization’s ability to navigate.

These sound similar. They’re not. Rebuilding trust is about the relationship between leaders and employees. Building navigational confidence is about the organization’s collective capability to move through uncertainty toward something worth reaching. The first is interpersonal. The second is directional.

The question employees are actually asking — beneath the surface demands for transparency and reassurance — isn’t “do I trust you?” It’s “do you know where we’re going, and can we get there?”

That question requires a different answer. Not reassurance. Navigation.

This is where AI strategy and culture have to connect — specifically and concretely, not abstractly. “We’re investing in AI” doesn’t move culture. “Here is what we’re building, here is why it creates durable competitive advantage, and here is what that means for the roles and capabilities we’ll need” — that starts to move culture. Because it gives people a destination to orient toward rather than a threat to brace against.

The narrative has to shift from loss to direction. Not everyone will make the transition. That’s the honest reality of disruption at this scale. The people who stay need something real to move toward. Abstract strategy doesn’t provide that. Specific navigation does.

Third Move: Make Employees Architects of AI Transformation

The deepest cultural shift available to leaders navigating AI disruption is also the most underutilized: bring the most-affected employees into the response itself.

Not as subjects of the transformation. As architects of it.

The instinct runs counter to how most organizations approach change. The AI strategy gets built in the C-suite or by a dedicated task force, usually populated with the most tech-forward employees. It gets rolled out to the broader organization as a communication exercise. And the people whose roles are most directly affected are the last to have meaningful input.

This is exactly backwards — both strategically and culturally.

Strategically, the employees closest to the operational work know where AI can genuinely augment what they do and where the friction will be. That intelligence is invaluable and largely untapped in most AI transformation efforts.

Culturally, the act of participation changes the psychological relationship to the change itself. People who helped shape the response don’t experience it the same way as people who received it. The anxiety doesn’t disappear, but agency does something to it. It transforms passive fear into active problem-solving.

Practical mechanisms matter here. AI working groups deliberately drawn from the most-affected roles, not just the most enthusiastic early adopters. Internal use-case development led by frontline employees with real operational context. Reskilling investments that are specific and credible — tied to actual roles the organization is building toward — rather than generic commitments. And visible wins attributed to the workforce, not to leadership or the technology itself.

The signal this sends — consistently, over time — is the one that actually shifts culture: we are navigating this together. You are not subject to this change. You are part of how we respond to it.

Chart 2: The Three Moves for Leading Through AI Disruption

Leading AI Culture Change Across Global Organizations

One aspect of AI culture transformation that most discussions skip deserves a clear flag: cultural transformation is not culturally neutral.

The three moves above have to be applied differently across different national and organizational cultures. Trust, transparency, institutional authority, and job security carry different weight in the US, UK, India, and China. What reads as honest and direct leadership in one context reads as alarming or disrespectful in another. The communication format that works in one office doesn’t land the same way in another country.

For executives leading global organizations through AI disruption, this isn’t a footnote. The frameworks have to be adapted, not just translated. And that adaptation requires local leaders with real authority and real context — not regional messengers for a centrally designed transformation.

This Disruption Is Different

The instinct to reach for a prior playbook is understandable. Leaders who navigated 2008, the internet era, and the mobile and cloud transitions built real muscle. The problem isn't the experience. It's that Agentic AI differs from every one of those disruptions across every dimension that matters for cultural leadership; where the threat originates, who absorbs it, how trust behaves, and what the workforce needs in response. The old frameworks weren't wrong. They were built for a different situation.

Chart 3: Why This Disruption Is Different

The Leadership This Moment Requires

Most executives were trained to lead from a position of confidence. Project clarity. Manage the narrative. Protect the organization from uncertainty until you have answers worth giving. That’s not bad leadership — it’s what most disruptions have required.

This one requires something different.

The Agentic AI disruption is happening in plain sight, at a pace that outstrips any organization's ability to get ahead of it, to a workforce that is simultaneously your greatest cultural asset and your most exposed. The old tools — the managed rollout, the optimistic town hall, the trust-rebuilding initiative — aren’t failing because leaders are applying them poorly. They’re failing because the situation has changed in a fundamental way.

What this moment requires is leaders willing to step into the uncertainty visibly, to trade the appearance of control for the credibility of honesty, and to build cultures that can navigate what isn’t fully known yet rather than cultures waiting for permission to engage with it.

That is harder than anything the standard change management curriculum prepared anyone for. It demands a tolerance for ambiguity that doesn’t come naturally to most people who reach the C-suite. And it requires a genuine belief that your people — including the ones most at risk — are capable of being part of the answer rather than just recipients of it.

Name it. Reframe it. Own it. That sequence won’t eliminate the anxiety. Nothing will. But it will give your people, and your culture, something the optimistic town hall never could: a reason to believe the navigation is real.

The leaders who get this right won't be remembered for having the best AI strategy. They'll be remembered for building organizations that could actually execute one — because the culture was ready, and the people trusted the navigation even when the destination wasn't fully in view.

That’s what leading through disruption has always meant, at its best. This moment just makes the demand more visible and the stakes harder to ignore.

Subscribe now

The views expressed in Uphoff on Media are entirely my own. They don’t represent the opinions of any company I’ve led, any board I’ve sat on, or any investor who’s had the pleasure of debating strategy with me over the years. If something I write here sounds brilliant, I’ll take full credit. If it turns out to be wrong, I was clearly misquoted by myself.

When Doug McMillon announced he was stepping down as CEO of Walmart after more than a decade at the helm, he didn’t blame the board. He didn’t cite personal reasons. He said something that should stop every senior executive in their tracks:

“With what’s happening with AI, I could start this next big set of transformations with AI, but I couldn’t finish.”

Read that again. The CEO of the largest company in the world; a leader who guided Walmart through the internet era, through supply chain collapse, through a pandemic, looked at what was coming and made a clear-eyed decision: the transformation horizon ahead didn’t match his own.

Around the same time, Coca-Cola CEO James Quincey said he could lead the company through the “pre-AI” phase, but that the next wave of “AI-driven growth” needed someone with fresh energy. Adobe CEO Shantanu Narayen, after eighteen years, framed his departure around the same idea: the next era of creativity is being “shaped by AI,” and new leadership should shape it.

Three CEOs. Three of the most recognized brands in the world. All saying a version of the same thing on their way out the door.

This isn’t a wave of high-profile executives burning out. It’s something more significant. It’s a cohort of experienced, accomplished leaders making a rational calculation that the technology disruption now underway is categorically different from anything they’ve navigated before — and deciding, honestly, that someone else should be at the wheel for the next leg of the journey.

That level of self-awareness deserves more credit than it’s getting. And it raises an urgent question for every other C-suite leader still in the chair:

What’s your honest answer to that same question?

The Data Behind the Signal

Before we get to frameworks, let’s establish the facts, because this isn’t anecdote. It’s a documented structural shift.

The numbers tell a story that goes well beyond a few high-profile departures.

CEO turnover hit record levels in 2025: a 16% spike in global departures from the prior year, more than 20% above the eight-year average. More than 1,200 U.S. CEOs left their posts in the first half of the year alone. Critically, this isn’t a story about underperformance. In the S&P 500, successions at top-performing companies jumped from 7% to 12%. Boards aren’t reacting to failure. They’re getting ahead of something.

The age data is where the AI signal becomes unmistakable: The average age of departing CEOs jumped from 55 to 63 in a single year. In July 2025, the average age of exiting CEOs hit 70.3, compared to 56.2 in the same month a year earlier. A retirement wave is underway, and AI is accelerating it. Boards aren’t just accepting these departures, in many cases they’re engineering them, actively skipping Gen X and moving directly to Millennial leaders for whom AI fluency is native, not acquired.

Here’s the wrinkle that most commentary is missing: not every experienced leader is heading for the exit. A recent survey found that 58% of Baby Boomers aged 62-plus said they are not currently considering retirement, a dramatic reversal from just a year prior, when only 11% said the same. The most common reason? They want to develop new skills. The best of this generation aren’t running from AI. They’re leaning into it. Which means the real divide isn’t generational. It’s between leaders who are engaging with the disruption honestly, and those who aren’t.

Which brings us to the real question: not whether AI is forcing a leadership reckoning—but how the leaders still in the chair should think about it.

A Disruption Unlike the Others

I’ve lived through several technology disruptions in operating roles; the internet’s first wave, the shift from print to digital, the rise of cloud and SaaS, the mobile platform shift. Each one was significant. Each one produced real winners and real casualties. None of them felt quite like this.

Here’s why this is different.

The internet disrupted business models and distribution. It reshuffled who owned the customer relationship and what information was worth paying for. But the core of what executives did — made decisions, managed people, read markets, allocated capital — remained fundamentally human. Technology changed the conditions. The decisions were still human.

I’ve watched this pattern play out firsthand. In the early days of the internet, I sat across from publishing executives who were genuinely unconcerned. “People will always want their morning newspaper. Magazines have loyal audiences that aren’t going anywhere”. The Internet platforms looked like novelties. They were not.

A few years later, while running The Hollywood Reporter, I was in meetings with studio executives exploring what digital distribution might mean for the filmed entertainment business. The response was remarkably similar. One executive turned my question about the still-nascent streaming model back on me: “Tony, theatrical box office is a $10 billion industry. Take a guess what DVD sales are. A $20 billion industry. We’re not spending much time thinking about streaming these days.” That was 2006. Today, Netflix is worth roughly two and a half times Walt Disney — the most powerful studio on the planet at the time of that conversation. The disruptor is worth more than twice the incumbent. The DVD business is a footnote.

The pattern across every major technology disruption is the same: the incumbents aren’t stupid. They’re rational, defending what’s working, discounting what’s early, and optimizing for the present. The problem is that the present has a shorter shelf life each time.

Agentic AI isn’t acting on the business. It’s acting on the function of decision-making itself. It’s entering the territory that senior executives have always owned: synthesis, judgment, resource allocation, workflow design, organizational structure.

That’s not a disruption to the business model. That’s a disruption to the role of the leader.

Consider what's now within reach:

CFO planning cycles that once required analyst teams for weeks can increasingly be run — at least in significant part — by AI agents.

Enterprise technology stacks that took CTOs and CIOs years to architect are being challenged by AI-native alternatives that bypass the complexity entirely.

Procurement functions that were built on human relationship and institutional knowledge are being tested against end-to-end automation.

Go-to-Market programs built over decades on personal relationships and market intuition are facing autonomous agents that evaluate, shortlist, and in some categories select vendors without a salesperson ever entering the picture.

None of this is fully realized yet. The question isn’t whether these shifts will happen. The capability is already here and moving fast. The question is whether the leaders in these roles are getting ahead of it or waiting for it to arrive. Because the executives who frame this as a threat to their function will lose that argument to the CFO who wants to cut headcount. And the executives who frame it as a transformation of their function’s value — from process management to strategic intelligence — are the ones who will define what these roles look like on the other side.

The EY analysis puts the productivity timeline plainly: the benefits of generative AI will likely arrive within three to five years, compared to multiple decades for the steam engine and roughly ten years for the computers. We are not in the early innings of a slow-moving transformation. We are in the compressed early innings of the fastest-moving general-purpose technology disruption in modern business history.

Morgan Stanley’s economists reviewed five major innovation waves; from the Industrial Revolution through the internet, and found that across every one of them, disruption ultimately complemented employment rather than eliminating it. History suggests we’ll adapt. But the historical record also suggests the transition is where leaders and organizations are made or broken.

The transition is now.

Subscribe now

The C-Suite Reckoning: Role by Role

The disruption isn’t uniform across the C-suite. Each function is facing a specific version of this challenge. Let me be direct about each one.

• The CEO. The pressure at the top is the most visible. The McMillon and Quincey moments are proof. But the more common failure mode isn’t the CEO who steps down honestly. It’s the CEO who stays, acknowledges the disruption publicly, and then delegates it to a Chief AI Officer or a transformation task force. Real AI transformation requires the CEO to own the conviction. Not every implementation detail, but the organizational belief that this is an enterprise-wide operating shift. Every time I’ve seen that conviction delegated, the transformation stalled. That’s more than a prediction. That’s a pattern I’ve watched repeat across every major technology shift for thirty-five years.

• The CFO. Traditional financial planning — scenario modeling, FP&A, cost forecasting — is going AI-native. The CFO who built their career on Excel mastery, quarterly reviews, and headcount-based cost structures is watching the foundation move. AI can run financial models in minutes that used to take FP&A teams weeks. The CFO who leads through this will be the one who reframes their role from financial architect to “capital allocation strategist”, deciding where human judgment is irreplaceable and where AI-driven analysis should drive the work.

• The CTO/CIO. Of all the C-suite roles, this one carries the sharpest irony. The executive whose entire mandate is technology leadership now faces a transformation that threatens to outrun the very playbook they built their career on. Enterprise technology stacks assembled over decades — the integrations, the vendor relationships, the architecture decisions — are being challenged by AI-native alternatives that bypass the complexity entirely. The CTO or CIO who leads through this isn't the one who defends the existing infrastructure investment. It's the one who can hold two things simultaneously: managing what the business runs on today while architecting what it needs to run on tomorrow. That's not a technology challenge. It's a leadership one.

• The CHRO. Workforce planning in an era when the CEO of Anthropic has publicly stated that AI could eliminate 50% of white-collar work within five years is genuinely uncharted territory. The CHRO who is still running traditional headcount plans and competency frameworks is already behind. The CHRO who will lead through this is building what I’d call a “dynamic workforce architecture”, a continuous model that maps AI capability against human work, identifies where augmentation creates leverage, and redesigns roles rather than eliminating them wholesale. The future is about Human + Digital Labor.

• The Chief Growth Officer / CMO. B2B marketing as practiced for the last decade is being structurally dismantled. Buyer journeys are being executed by AI agents rather than human researchers. The content, events, and paid media programs that generated pipeline in 2020 are producing diminishing returns. The CMO who leads through this transition will stop optimizing existing channels and start designing for a world where the first buyer in the room may not be human.

• The Chief Procurement Officer. This may be the function most immediately in the crosshairs. Agentic AI is coming for sourcing, vendor evaluation, contract management, and supplier intelligence with a speed and scale that no traditional procurement team can match. The CPO who leads through this isn’t managing a cost center. They’re running an intelligence function, one where the value isn’t in the process they oversee, but in the strategic judgment they bring to decisions that AI can inform but not make.

Subscribe now

Three Archetypes for Leading Through This

I’ve spent time thinking about how C-suite leaders are actually responding. Not how they’re describing it in the press release, but how the work is actually unfolding. What I see are three archetypes. Each has a different job to do.

The Transformer

This is the leader who has the runway — in terms of time horizon in the role — and the conviction to lead through the full AI transformation cycle. The imperative for the Transformer is to stop treating AI as a department initiative and start treating it as an operating system for the enterprise.

What that means in practice: every function gets reviewed through the lens of what AI can do that humans were doing, what humans must do that AI cannot yet do, and where the hybrid model creates genuine competitive advantage. That’s not a technology project. It’s a strategic redesign, and it has to be led from the top.

The Transformer also has to be willing to do something most executives avoid: change the incentive structures before the transformation demands it. AI adoption stalls when the people responsible for it are also the people most threatened by it. If the CHRO’s success metrics are still tied to headcount growth, you will not get genuine AI workforce transformation. If the CPO is measured on cost savings from vendor relationships, you won’t get a procurement function redesigned for intelligence-led sourcing.

The Transformer changes the metrics first.

The Sequencer

This is the leader who is honest — privately, at minimum — that they won’t finish the journey. They have enough runway to make real moves, but they can see the horizon. The Sequencer’s imperative is not transformation, it’s architecture.

What does a Sequencer do? They build the capability scaffold that gives the next leader a genuine running start. That means: driving AI literacy into the organization at every level. Building the data infrastructure that AI-driven operations require. Making the succession bet: identifying and developing the leaders who have both the operating experience and the AI fluency to take the baton.

The worst outcome for a Sequencer is the one I’ve seen too often in technology transitions: a leader who neither transforms nor prepares. Who manages inertia, makes incremental moves, protects existing structures, and hands the next CEO a business that is further behind than it looks. That’s not a neutral outcome. It’s an actively harmful one.

The Sequencer who does their job well is underappreciated in real time and overappreciated in hindsight. Do it anyway.

The Advisor

This is the leader who makes the call that McMillon made. Honestly, on their own terms. They step out of the operating role, but they don’t disappear. They move into the work that their experience actually makes them irreplaceable for: board roles, advisory relationships, mentoring the next generation of operators.

I’ll be transparent: this is the archetype I’m living right now. I recently stepped out of a CEO role, one where we spent two-plus years transforming a B2B demand generation business into a digital marketing services company built on a Demand-as-a-Service model. We started the AI transformation. We could see what was coming. And I made the same calculation McMillon described, at a far smaller scale. The next leg of the journey was better led by someone whose horizon matched it. That recognition is what led me to launch Uphoff Advisory and shift into board and advisory work — where thirty-five years of operating experience turns out to be exactly what organizations navigating this transition need most.

The Advisor role isn’t a retirement. It’s a repositioning.

The Honest Question

Let me end where I started.

McMillon said: ”I could start it, but I couldn’t finish it.”

That is a remarkable sentence. It requires both confidence; the conviction that you understand what needs to happen — and humility — the honesty that the journey ahead isn’t yours to complete.

Most leaders in the C-suite right now are not asking themselves that question. They’re managing the present. They’re responding to quarterly pressures, board expectations, and the urgent over the important. The AI disruption is real and they know it, but it hasn’t forced a clear-eyed personal reckoning yet.

It will.

The leaders who will navigate this best, whether as Transformers, Sequencers, or Advisors, are the ones who ask the question now, honestly, and then act on the answer with the same rigor they’d apply to any other strategic decision.

The Transformer commits the enterprise to a full operating redesign and changes the incentives to match.

The Sequencer builds the scaffold, develops the successor, and makes the architecture decisions that matter most.

The Advisor doesn't exit. They redeploy, bringing years of pattern recognition into the boardroom, into advisory relationships, and into the work of guiding operators navigating a disruption they've already lived through once.

None of these are failure modes. The only failure mode is the one that looks like leadership but is actually avoidance: staying in the chair, acknowledging the disruption in every earnings call, and changing nothing that actually matters.

Subscribe now

The views expressed in Uphoff on Media are entirely my own. They don’t represent the opinions of any company I’ve led, any board I’ve sat on, or any investor who’s had the pleasure of debating strategy with me over the years. If something I write here sounds brilliant, I’ll take full credit. If it turns out to be wrong, I was clearly misquoted by myself.

By Tony Uphoff · Uphoff on Media · 2026

In my ongoing conversations with B2B media, marketing, and tech leaders, the subject of Vibe Coding keeps coming up.

I’ve been shown remarkable engineering productivity gains — 10X+ — by multiple, experienced software developers. I’ve had several conversations with Enterprise Tech leaders who chuckle and suggest the hype is overblown and that Vibe Coding won’t be challenging Enterprise Software anytime soon…in my experience a sure sign that it likely will! And my discussions with media and marketing leaders suggest there is a ton of interest, some experimentation, and — most critically — a need for frameworks to understand the technology and how to use it.

This post is that framework. Consider it a primer, a backgrounder, and a practical map for non-technical executives (like myself), operators, and investors navigating one of the more consequential shifts in how software gets built.

I. WHAT IS VIBE CODING?

Vibe Coding was coined by Andrej Karpathy, former Tesla AI director and OpenAI co-founder, in February 2025. His description was characteristically precise: you describe what you want in natural language, and AI writes the code. You stop thinking about syntax. You stop debugging semicolons. You stop waiting for a developer to have a sprint slot available.

The human becomes the director. The AI becomes the builder. The interface between intention and output is a conversation.

It’s not that the code writes itself. It’s that you describe the outcome, and the code appears.

This is meaningfully different from autocomplete tools or basic code suggestions. Vibe Coding generates full working codebases from plain English prompts, iterates through back-and-forth conversation, and handles everything from the front-end interface to the back-end logic, without the operator needing to understand how any of it works at the implementation level.

The tools driving this are now mature and widely available:

• Cursor: the leading AI-native code editor, built for Vibe Coding workflows

• Bolt, Lovable, and Replit Agent: browser-based tools for full-stack app generation with no setup required

• GitHub Copilot: Microsoft’s enterprise-grade AI coding layer, now deeply embedded in developer workflows

• Claude Code: Anthropic’s command-line agent for complex, multi-file codebases

• Windsurf: A newer entrant gaining traction with enterprise development teams

All of these tools are powered by large language models (LLMs) underneath. The same class of AI that powers conversational interfaces like ChatGPT and Claude. What the Vibe Coding tools add is the scaffolding, context management, and workflow design that makes LLMs practically useful for software development.

Figure 1: The Anatomy of Vibe Coding: from intent to output

II. THE WORKFLOW: HOW IT ACTUALLY WORKS

Understanding the Vibe Coding workflow is essential to evaluating whether and how it applies to your organization. The process follows four steps:

Step 1: Describe in Plain English

The operator; a marketing manager, a media product director, a solutions engineer, describes what they need in natural language. “Build me a lead capture page that connects to Salesforce and sends a confirmation email.” “Create a dashboard that shows subscriber engagement by content type.” “Build an event registration flow with payment processing.”

Step 2: AI Generates the Full Codebase

The Vibe Coding tool interprets the prompt and generates a working application: HTML, CSS, JavaScript, back-end logic, database schema, API connections. It’s not a mockup or a template. It’s functional code.

Step 3: Iterate via Conversation

This is the step most people underestimate. Vibe Coding is not a one-shot prompt. It’s a dialogue. “Make the button gold instead of blue.” “Add a dropdown for industry.” “The form isn’t submitting, fix it.” The AI debugs, adjusts, and refines in real time, maintaining context across the conversation.

Step 4: Ship

The finished application is then deployed to a web host, a platform, a CRM integration. What used to require a developer, a sprint, a QA cycle, and a deployment process now takes hours.

The development backlog as a structural constraint is being dismantled. Not reduced. Dismantled.

III. WHERE IT’S LANDING: B2B APPLICATIONS

Vibe Coding is not a consumer technology story. Its most significant near-term impact is in B2B. Specifically in the three sectors that Uphoff on Media covers most closely: media, marketing, and technology. The applications are concrete, the productivity gains are measurable, and the implications for how these organizations staff, build, and compete are substantial.

The most instructive examples often come from direct experimentation. I’ve been experimenting firsthand. Using Tasklet; an agentic AI platform that automates business workflows through plain English, I built a client tracker for Uphoff Advisory and a research tool that surfaces data on Agentic AI for future posts. No developer. No code. Just a description of what I needed and an AI that executed it. Creating the two apps took a total of 5 minutes. That’s not quite Vibe Coding in the technical sense — it’s something arguably more immediately useful for business operators: agentic workflow automation. The line between the two is blurring fast.

Figure 2: Vibe Coding Applications Across B2B Media, Marketing & Technology

Subscribe now

B2B Media

For media companies, the most immediate impact is in the elimination of the dependency on engineering resources for tools that editorial, audience development, and monetization teams need. Newsletter templates, audience analytics dashboards, event microsites, advertiser-facing portals. All of these have historically required developer time that editorial teams rarely had access to. Vibe Coding changes the equation: the people closest to the audience and the advertisers now build the tools they need, on the timeline they need them.

The deeper implication: media organizations that have always competed on content quality now have the ability to compete on product velocity as well. That’s a new capability, and it arrives without a significant increase in headcount.

B2B Marketing

In B2B marketing, the impact is most acute in the gap between marketing’s ambitions and engineering’s availability. ABM campaign landing pages, lead scoring applications, content performance dashboards, ICP segmentation tools. These are all things that marketing teams have wanted for years and have had to either deprioritize or outsource. Vibe Coding puts these capabilities directly in the hands of demand generation managers, RevOps leads, and marketing ops practitioners.

The signal to watch: in the companies I’ve spoken with that have begun experimenting with Vibe Coding in their marketing functions, the productivity gains are not incremental. They are order-of-magnitude. A campaign that required three weeks of engineering time and agency involvement now ships in a day.

B2B Technology

For technology companies, Vibe Coding is simultaneously the most significant opportunity and the most complex challenge. On the opportunity side: internal admin tools, API integration prototypes, product demo environments, and MVP development are all areas where Vibe Coding is delivering demonstrable results. Solutions engineers are building custom proof-of-concept integrations in hours. Product managers are prototyping new features without waiting for sprint allocation.

The complexity: for companies whose core product is software, Vibe Coding raises fundamental questions about code quality, security, maintainability, and ownership. The engineers who chuckle at the hype are not wrong to raise these concerns. They are, however, at risk of underestimating the pace of change.

IV. THE PRODUCTIVITY REALITY

The data on Vibe Coding productivity is early but directionally consistent:

• GitHub reports: Copilot users complete coding tasks up to 55% faster than non-users

• McKinsey research finds: AI-assisted developers complete tasks 35–45% more quickly

• Anecdotal operator data: including from conversations in my own advisory network — suggests that experienced developers using Vibe Coding tools are achieving 5–10X productivity multipliers on certain task types

That last point is important context. The most dramatic productivity gains are not coming from non-technical users suddenly building enterprise software. They are coming from experienced developers who now spend less time on boilerplate, scaffolding, and syntax. And more time on architecture, judgment, and quality. The 10X+ gains I referenced in the opening of this post came from two senior developers with decades of experience. Vibe Coding amplified their expertise; it did not replace their judgment.

The 10X gains I’ve witnessed came from senior developers with 35 years of experience each. Vibe Coding amplified their expertise. It did not replace their judgment.

V. THE LEGITIMATE CONCERNS

The Enterprise Tech leaders who are skeptical of Vibe Coding are not simply behind the curve. Their concerns are substantive:

• Technical debt and code quality: AI-generated code can be structurally sound but architecturally fragile. Code that works is not the same as code that scales, that can be maintained, or that follows security best practices.

• Security exposure: Non-technical operators building applications that handle customer data, payment information, or system integrations create real security risk if the code is not reviewed by engineers with security expertise.

• IP and ownership ambiguity: The legal framework around AI-generated code; who owns it, how it’s licensed, what claims can be made, is still being established.

• The Dunning-Kruger risk: The most dangerous Vibe Coding practitioner is the one who doesn’t know what they don’t know. Building something that appears to work and shipping it without understanding its failure modes is a real organizational risk.

None of these concerns are arguments against Vibe Coding. They are arguments for deploying it with judgment. Which, as it happens, is what every consequential technology shift has required.

VI. THE FRAMEWORK: HOW B2B LEADERS SHOULD THINK ABOUT THIS

Based on conversations with operators across B2B media, marketing, and technology, I’d suggest the following framework for evaluating and deploying Vibe Coding:

• Inventory Your Queue First

Before experimenting with Vibe Coding tools, map the engineering backlog items that belong to non-engineering functions: marketing’s dashboard requests, media’s editorial tools, the internal ops workflows that never make it into a sprint. That backlog is your starting point. It represents the highest-value, lowest-risk surface for Vibe Coding experimentation.

• Start with Internal Tools, Not Customer-Facing Products

Internal dashboards, reporting tools, and workflow automation are the right first applications for Vibe Coding in most B2B organizations. The stakes of failure are lower, the learning curve can be absorbed without customer impact, and the productivity gains are immediately visible.

• Keep Engineers in the Review Loop

Vibe Coding does not make engineering judgment obsolete. It changes where that judgment is applied. For any application that touches customer data, payment systems, or core business processes, engineer review of the output is not optional. It is the risk management layer that makes Vibe Coding safe to deploy at scale.

• Treat Vibe Coding as a Skill, Not a Tool

The operators in my network who are getting the most from Vibe Coding are not simply running prompts. They are learning how to structure requests, how to iterate effectively, how to recognize when the AI has gone off track, and how to maintain context across a complex build. This is a learnable skill. Organizations that invest in developing it will have a meaningful advantage over those that treat it as a plug-and-play solution.

• Watch the Enterprise Software Implications

The Enterprise Tech leaders who suggest Vibe Coding won’t challenge SAP, Oracle or Salesforce “anytime soon” are probably right on the one-to-two-year horizon. On the three-to-five-year horizon, the trajectory is less clear. Vibe Coding is already enabling the rapid development of lightweight alternatives to expensive, complex enterprise tools. The category of “good enough at a fraction of the cost” has disrupted enterprise software before. It will again.

THE BOTTOM LINE

Vibe Coding is not hype. It is also not magic. It is a meaningful shift in the accessibility of software development. One that is already producing measurable productivity gains, enabling new categories of operators to build tools they couldn’t before, and beginning to reshape how B2B companies think about the relationship between their business teams and their engineering resources.

For B2B media leaders, the opportunity is in product velocity and editorial tool ownership. For B2B marketing leaders, the opportunity is in closing the gap between what you want to build and what you’re able to ship. For B2B technology leaders, the opportunity is in developer productivity and the acceleration of prototyping and internal tooling.

The risk in all three cases is the same: deploying a powerful capability without the judgment infrastructure to use it well.

Vibe Coding doesn’t replace engineers. It eliminates the queue. Whether your organization knows the difference will matter.

The pattern is familiar.

The internet didn’t eliminate publishing. It eliminated the gatekeepers of distribution.

SaaS didn’t eliminate IT. It eliminated the bottleneck of infrastructure.

Vibe Coding won’t eliminate engineering. It will eliminate the monopoly that engineers have historically held on who gets to build.

For the executives, operators, and investors reading this: the question isn’t whether Vibe Coding is real. It’s whether your team is already using it, and whether you know.

Subscribe now

The views expressed in Uphoff on Media are entirely my own. They don’t represent the opinions of any company I’ve led, any board I’ve sat on, or any investor who’s had the pleasure of debating strategy with me over the years. If something I write here sounds brilliant, I’ll take full credit. If it turns out to be wrong, I was clearly misquoted by myself.

We learn by hearing stories — especially ones told with authenticity, relevance, and a bit of humor. "Leadership Lessons Learned the Hard Way" is a regular feature on Uphoff on Media, drawing from my own experience: the Hollywood producer who hit me with the classic "you'll never work in this town again," the time my go-to sign-off "Best, Tony" autocorrected to "Beast, Tiny." Real situations. Hard-won lessons. For the third installment, I'm sharing one taught to me by a customer. One I've never forgotten.

I was a young sales rep at the B2B publication EE Times, freshly transferred to Silicon Valley. Big opportunity. Big territory. And assigned to one of the most powerful ad agencies in the Valley. A firm that represented household names in tech and hadn’t done business with EE Times in years.

The VP Media Director was legendary. She could make or break media brands. And careers. Getting a meeting with her was itself a milestone.

So I did what young, hungry sales reps do. I called. I left voicemails. Detailed, articulate voicemails laying out our value proposition and requesting a meeting. Every week. For six months.

Crickets.

Until one evening at 6:00pm, my office phone rang. I picked up. After I said hello, the voice on the other end said:

“You’re not going to go away, are you?”

We met. She laid out exactly how she saw the market, what her clients needed, and how she evaluated media brands, including mine. It was a masterclass. Over the next couple of months we met several times, leading to EE Times being invited to participate in a media day for one of their biggest clients.

This was enormous. My publisher, editor, and associate publisher were flying in from New York for a Friday meeting. Careers were watching.

On Wednesday, she called and asked me to bring the presentation over. I told her it was still a work in progress. She didn’t care.

I walked her through a printed copy of 50 slides. She sat with her arms folded the entire time. When I finished, she was quiet for a moment. Then she nodded and said:

“This is good. You’re not telling me I’ve been stupid for not buying EE Times. What you’re telling me is that the market has changed, your editorial has captured that change, your readership reflects it, and now I should consider it.”

I blushed beet red.

Because that’s not what the presentation said at all. It was a competitor takedown dressed up as a value proposition. Slide after slide subtly — and not so subtly — making the case that anyone who hadn’t been buying EE Times must be clueless.

She had just described the presentation I needed to give. Not the one I had.

I gathered up the presentation, thanked her, drove back to the office, and spent the next three days rebuilding everything with my team. We threw out most of what we’d built. We reframed every slide around a single organizing idea: the market had moved, EE Times had moved with it, and here was the evidence. No competitive takedowns. No implicit verdicts. Just a clear-eyed view of a changed landscape and where we fit in it.

Friday came.

The room was full. Senior agency team, client marketing leadership, my New York executives. The kind of meeting where everything either opens up or closes down.

We gave the new presentation.

It worked.

The result was the largest single sale of my career to that point, and one of the largest single sales EE Times had recorded. The agency became a real partner. The client relationship grew.

But the outcome I didn’t anticipate was what happened next with the VP Media Director herself.

She became a mentor.

Subscribe now

Over the years that followed, she recommended me for leadership positions. She saw something in that three-day rebuild, in the willingness to throw out work I was proud of and start over because someone wiser showed me the truth, that told her something about how I’d lead. She invested in my career the way she’d invested in that Wednesday afternoon conversation. Quietly. Decisively. Without making a big deal of it.

I would not have had the management career I had without her mentorship.

Here’s what she taught me, and what I’ve carried across five CEO roles since:

When you’re trying to change someone’s mind, you don’t win by making them feel wrong. You win by showing them the world has changed.

There’s a version of persuasion that’s really just an indictment. It says: here’s all the evidence that your current approach is mistaken, your current vendors are inferior, your current thinking is outdated. It feels like a strong argument. It’s actually a trap, because no one changes their mind when they feel attacked. They defend. They dig in.

And they remember how you made them feel long after they’ve forgotten your slides.

The better strategy, the strategy she modeled for me before I even knew I needed it, is to lead with the landscape, not the verdict. The market shifted. New signals emerged. A different approach may now make sense. That framing gives the buyer somewhere to go. It honors the intelligence of the decisions that got them here, while opening a door to something new.

This applies well beyond sales. I’ve used this frame in boardrooms, turnaround conversations, investor updates, one-on-ones and leadership team offsites. Any time you’re asking someone to change course; a strategy, a vendor, a belief, the question isn’t how do I prove them wrong. It’s how do I show them the landscape has moved in a way that makes a new direction logical, even obvious?

One more thing worth naming: she didn’t have to do what she did on that Wednesday afternoon.

She could have let me walk into Friday’s meeting and fail. Instead she called, sat with her arms folded through 50 bad slides, and handed me the insight that changed everything. That’s its own leadership lesson. The most powerful people in any room often got there because they bring others along rather than let them stumble.

She did that for me. More than once.

I’ve tried to pay that forward ever since.

Missed the first two? Start here: “You’ll Never Work in This Town Again” and here: “Beast Tiny. What Autocorrect Taught me About Leading at the Speed of Chaos”

Subscribe now

The views expressed in Uphoff on Media are entirely my own. They don’t represent the opinions of any company I’ve led, any board I’ve sat on, or any investor who’s had the pleasure of debating strategy with me over the years. If something I write here sounds brilliant, I’ll take full credit. If it turns out to be wrong, I was clearly misquoted by myself.

A few years ago, I started getting calls from young people in their junior year of college. Friends of our daughter, classmates, people looking for a perspective on what to do next. I was glad to take them. I’ve been in and around B2B media, marketing, and technology for more than 35 years. If I could save someone a few wrong turns, I was happy to help.

One of those conversations has stayed with me. A young woman — a high school valedictorian, Ivy League junior, internships at Google and NBC, job offers from both — reached out for some perspective. She was leaning toward NBC. When I asked why, she said her parents felt it was the safer choice. They knew NBC. It was familiar. Recognizable.

I told her: “You’ll do well no matter where you go. What this really comes down to is whether you want to be part of the past or the future.”

She thought about it. Then she went into consulting.

I’ve thought about that outcome more than she probably has. She didn’t choose the past or the future. She chose optionality. She intuitively understood something that took me years to be able to articulate: the institution itself was the risk. At 21, without having read a single business book about it, she made an infinite game move.

That instinct is rarer than it should be. And in 2026, the cost of not having it has never been higher.

The numbers are not ambiguous

Let’s start with the data, because this post is not about feelings. It’s about facts, and the facts in B2B media, marketing, and technology are stark.

• 17,000+ entertainment and media jobs cut in 2025: An increase of 18% from the year before

  (Challenger, Gray & Christmas)

• 54,000 advertising, PR, and related services jobs lost in a single year through May 2025

  (Bureau of Labor Statistics)

• 14,298 average annual media job cuts since 2018: Nearly double the 7,305 annual average from 2010–2017

• Informa TechTarget, formed in December 2024 with 2,000 employees, had shed roughly 500 of them by October 2025.

On the agency side, the story is the same. Forrester analysis puts agency headcount down 8% in 2025. The Omnicom-IPG merger alone is expected to generate $750 million in cost savings, and labor costs always lead the balance sheet. WPP, Dentsu, IPG: thousands of roles gone, with more to follow.

The jobs disappearing fastest are the ones that taught the craft. Junior copywriter. Media planner. Demand gen coordinator. Research analyst. The entry-level roles that once served as the apprenticeship layer of this industry; where you learned how audiences work, how campaigns are built, how buyers actually behave, are being automated out of existence first. Anthropic’s own research shows hiring of younger workers has slowed roughly 14% in the most AI-exposed occupations compared to 2022.

If you are early in your career in this industry, or advising someone who is, this is not a cycle. It is a structural shift. The map you were handed is of a city that no longer exists.

A book changed how I think about all of this

I read Simon Sinek’s “The Infinite Game” in 2019. To this day, no single book has more fundamentally shaped — and reshaped — my philosophy about work and career. The thesis hit me hard.

We have been brought up in business to apply finite game theory to infinite games. Finite games have a fixed endpoint, agreed-upon rules, and a winner. Infinite games have no finish line. The only goal is to keep playing.

Careers are infinite games. We have been playing them like finite ones.

The ultimate goal of the career game is to keep on playing. Not win a single position, title, or business deal.

“The Infinite Game” became my most gifted book for the next three years. I revisit it often. Because every time I look at the career decisions I’ve watched people make — and the ones I’ve made myself — I see the same pattern. Someone optimizing to win the quarter, land the title, close the deal. And in doing so, weakening their position in the game that actually matters.

The valedictorian who went into consulting wasn’t running from a decision. She was, instinctively, playing the longer game. The parents who steered toward NBC were giving advice from inside their own era. Advice that was entirely correct for a media landscape that no longer exists. That’s not a failure of love or intelligence. It’s what happens when you use last decade’s map to navigate this decade’s terrain.

Subscribe now

What a career in this industry actually is

Thirty-five years. Five CEO roles. This is the framing that has held up through all of it:

A career in B2B media, marketing, and technology is not a series of jobs. It is a continuously developed capability to understand how buyers find, trust, and act on information, and to create value at that intersection across media, marketing, and technology.

That’s it. Everything else: the titles, the companies, the platforms, the tech stacks, is scaffolding. The scaffolding changes. The function doesn’t.

The companies that collapsed in this industry were the ones that forgot what function they actually served. They started optimizing for the advertiser or the algorithm instead of the buyer. The careers that stalled were the ones that attached too tightly to a form; a specific medium, a specific platform, a specific role, rather than the underlying function.

Agentic AI is not changing the function. It is reorganizing everything around it. The question for anyone building a career in this industry right now is not: “Will my job exist?” It is: “Am I building toward the function or toward the form?”

A framework for three stages

The infinite game plays out differently depending on where you are in it. Here is how I think about it across three stages. Use them as a mirror, not as a map

Stage 1 · Years 0–10

Get close to the work

The goal is not the credential. It is closeness to the actual work of understanding buyers, building audiences, and creating demand.

The danger at this stage is chasing the brand-name employer, the prestigious title, the defined role, at exactly the moment those institutions are most unstable. Get as close as possible to the actual work. Learn the underlying physics. Not the specific machine.

Diagnostic questions:

1. Can you describe, in plain language, how a B2B buyer actually makes a purchase decision…not in theory, but in the specific market you work in?

2. If your current employer disappeared tomorrow, what capability would you carry out the door, and how would you describe it to the next person?

3. Are you learning from people doing the work, or from people managing those who do?

• The honest answer is the former. Managers teach you how the institution works. Practitioners teach you how the work works. Only one of those travels with you.

4. Is your current role building a skill, or executing a process someone else designed?

5. When did you last do something at work that made you uncomfortable because you weren’t sure you could do it?

Drift signal: You are optimizing for the brand name on your resume rather than what you are learning inside the building. The institution feels like the asset. It isn’t. What you can do when you leave is the asset.

Stage 2 · Years 10–20

Build something portable

The goal is not the next title. It is a reputation, a network, and a point of view that exist independently of any single institution.

This is where Sinek’s insight lands hardest. Mid-career professionals in B2B media, marketing and technology have usually been playing a finite game, competing for the next title, the next company, the next comp band. The infinite reframe asks: what are you building that compounds?

Diagnostic questions:

1. If someone in your industry were asked about you by name, what would they say you are known for? Not your title, not your company, but you?

2. In the last 12 months, have you built any relationships that exist outside your current employer’s org chart?

3. Do you have a point of view about your industry that you could defend in a room full of smart people who disagree with you?

4. If your company restructured tomorrow and your role disappeared, how long would it take you to land your next opportunity, and what would that process look like?

5. Are the decisions you are making now — projects, relationships, visibility — compounding, or are they just getting you through the quarter?

Drift signal: Every career move for the past five years has been toward a better title or higher comp, with no conscious thought about whether it built something portable. The ladder is the strategy. It isn’t enough.

Stage 3 · Years 20+

Recognize what you actually have

The goal is not survival. It is accurate inventory. Most senior operators in this industry underestimate the scarcity of what they have built.

Late-career operators in B2B media, marketing, and technology have something genuinely scarce: they understand how buyers actually behave, how deals actually get made, how organizations actually change. Not in theory, but from having been inside the room when it happened. Agentic AI can automate the process of demand generation. It cannot replicate the judgment about which markets matter, and why, and when.

Diagnostic questions:

1. What do you know about how buyers, markets, or organizations actually behave that cannot be found in a report, a course, or an AI prompt?

2. Who calls you when they have a hard decision to make, and why do they call you instead of someone else?

3. Are you framing your experience in terms the next generation of decision-makers can understand and act on, or in terms that only made sense in the media and marketing economy of 15 years ago?

4. What would you build, advise, or create if you optimized for the next 10 years of the game rather than the next role?

5. Are you giving advice from who you are now, or from who you were when you were most successful?

Drift signal: You are describing your experience in terms of institutions and titles rather than judgment and outcomes. The instinct is to protect the identity you built. The opportunity is to deploy what is behind it.

The framework above is also available as a standalone reference. Download or save this for reference, and share it with someone who needs the framework.

The real question isn’t about AI

Every version of this conversation eventually arrives at AI. And the conventional answer is always the same: learn to use it, stay adaptable, embrace change. That advice is true and useless in equal measure.

The more interesting question is the one the valedictorian answered without knowing she was answering it. Not: “What tool do I need to learn?” But: “What am I building that is valuable with or without any institution behind it?”

In B2B media, marketing, and technology, the answer to that question has been the same for decades, even as every platform and format around it has changed. The buyers are out there. They need to find solutions they can trust. Someone has to build the bridge. The function has always been to be part of that bridge; credibly, durably, in a way that compounds over time.

The response to my last piece made clear that the economic pressure is real and widely felt.

The question I keep getting, in various forms, is the same: “Given all of this, what do I do with my career?” This is my best current answer.

A career in B2B media, marketing, and technology is not a series of jobs. It is a continuously developed capability to understand how buyers find, trust, and act on information — and to create value at that intersection. The function has been the same for decades. The scaffolding around it; the titles, the platforms, the institutions, the tech stacks, has changed repeatedly, and is changing again now, faster than at any point in my career.

The map has changed. The territory hasn’t.

Play the infinite game.

One question I'll leave you with — and one I'd genuinely like to hear your answer to: which stage of this framework are you in, and are you building toward the function or toward the form?

Subscribe now

The views expressed in Uphoff on Media are entirely my own. They don’t represent the opinions of any company I’ve led, any board I’ve sat on, or any investor who’s had the pleasure of debating strategy with me over the years. If something I write here sounds brilliant, I’ll take full credit. If it turns out to be wrong, I was clearly misquoted by myself.

This post grew out of conversations with CEOs running B2B media companies, investment bankers evaluating them, and board members facing hard questions about the industry's future. Every conversation circled back to the same unspoken truth: the digital advertising model these businesses migrated to was never a real replacement for what they left behind. It was a discount that became a floor.

I've spent 35 years inside these businesses. I know what the numbers looked like when they worked — and what it actually takes to transform them. What follows is an honest account of how the industry got here and where it goes next.

When Print Was a License to Print Money

In the late 1990s and early 2000s, a B2B trade magazine serving a specialized vertical — information technology, process manufacturing, electrical contracting, food processing — could command a $107 CPM for a full-page, four-color advertisement. A specialty title with 60,000 audited, qualified subscribers could price a full page at $8,000 gross, a CPM of approximately $133. After a standard agency discount, the net still landed around $107.

Today, the same professional audience, reached through programmatic digital display generates a CPM of $5 to $7.

That’s a 93%–95% erosion in effective pricing power. For the same eyeballs. Over 25 years.

Subscribe now

Speaking from Experience: In 1999 I was Publisher of InformationWeek Magazine, one of the largest B2B brands in the world. That year we did over 7,000 ad pages at $25,000 per page. Read that again…seven thousand ad pages. In a single year. That’s $175 million in print advertising revenue. From one magazine. In one year.

The premium wasn’t arbitrary. It reflected something real: Scarcity.

If you wanted to reliably reach a technology decision-maker, a licensed electrical contractor, or a process engineer at scale, there was essentially one way to do it; buy ad space in the trade publication they subscribed to, held in their hands, and read professionally.

The math was compelling on both sides. The high-demographic subscriber who received the magazine and engaged with it over time wasn't just an audience member. They were a demonstrably valuable buyer. Pass-along readership multiplied the effective reach beyond the core circulation. Advertisers understood this intuitively even before there was data to prove it.

By the early 2000s the numbers started to tell a different story. One nobody wanted to read. The average B2B magazine still had healthy CPM’s. Same brand. Same audience. Online CPM was 75% lower. Publishers looked at that gap and called it temporary. The investment growth phase. An early-market anomaly that scale and time would correct. They were wrong. The discount wasn’t a phase. It was the new floor, and it only went lower from there.

The Collapse: A Timeline

Late 1990s – 2002: Peak Print Era: B2B trade magazines commanding $80–$133 CPMs for qualified vertical audiences. Advertising revenue funds deep editorial operations. The scarcity moat is real. Nothing else reaches this buyer.

2000–2005: The Original Digital Discount: Publishers launch websites and price banner inventory at $15–$20 CPM to attract online advertisers — a 75–85% discount to their own print rates. The logic: build traffic, premium pricing will follow. It never does.

2005–2013: The Lead Gen Insurgency: B2B publishers begin offering something print never could: a verified contact with demonstrated content intent. TechTarget pioneers the model at scale; IDG, Ziff Davis, and UBM TechWeb follow. Brand advertising budgets migrate to guaranteed lead programs, starving display advertising of its most committed buyers. A parallel race to the bottom hits as low-cost lead providers flood the market, compressing CPLs across the entire category.

2007–2012: Programmatic Goes Mainstream: Real-time bidding exchanges launch at scale. Programmatic platforms aggregate supply from millions of publishers. B2B display CPMs fall below $5 on open exchanges. Publishers lose control of their pricing power. Permanently.

2012–2016: Mobile and Social Compound the Supply Explosion: Facebook and Google vacuum up B2B advertising budgets with precision targeting and infinite scale. LinkedIn captures the professional audience CPM premium that publishers once owned. B2B trade brands are left competing for scraps.

2015–2020: The Quality Counter-Movement: High-quality content syndication providers build genuine first-party data marketplaces. Owned, permission-based audiences with verified professional identities, clean compliance frameworks, and behavioral intent signals. The model proves something the skeptics got wrong: quality and scale can coexist. It also draws a sharp line between publishers who protect their audiences and those who have been quietly burning them for short-term revenues.

2017–2022: The Roll-Up Era: PE-backed consolidators acquire fragmented B2B media assets on the thesis that scale produces margin. Debt loads rise. Digital CPMs plateau at $5–$7. Events become the revenue lifeline. The problem is deferred, not solved.

2023–2025: AI Disrupts Search Discovery: Generative AI begins disintermediating organic search traffic, the primary driver of publisher page views. CPM floors drop. Traffic volumes erode. Both revenue pillars collapse at once.

2026 →: The Agentic Inflection: Autonomous AI agents increasingly handle B2B research, vendor discovery, and procurement intelligence, the functions B2B media brands built their entire editorial identity around. The CPM question is now secondary. The real question is whether the audience relationship survives at all.

The Underlying Problem Nobody Wants to Say

The CPM collapse isn’t primarily a technology story. It’s a story about what publishers actually sell.

For decades, B2B trade magazines sold access to scarce audiences. That scarcity was real. There was no other reliable way to reach an IT professional, licensed electrician or a food plant manager at professional scale. Publishers monetized that scarcity at premium rates.

The internet didn’t destroy the audience. It destroyed the scarcity.

When every professional became findable through LinkedIn, searchable through Google, and targetable through guaranteed lead generation and programmatic platforms, the publisher’s core value proposition: “we have access to people you can’t reach otherwise”, quietly evaporated. And most publishers didn’t notice until the revenue was already gone.

What remains is editorial credibility, community trust, and first-party data value. These are real assets. But they require a completely different operating model, revenue architecture, and executive mindset to monetize. Most legacy B2B publishers have never built for any of those things, because for decades, they never had to.

Subscribe now

The Lead Gen Trap That Nobody Talks About

There’s a third force in the CPM collapse that gets less attention than programmatic or mobile, but may have done more structural damage to B2B publishing than either: The rise of guaranteed digital lead generation.

The pitch was irresistible. Stop paying for impressions that might reach someone vaguely relevant. Start paying only for leads you can put directly into your CRM. When the average cost per lead settled around $43 for a verified, intent-qualified contact, the case for paying $40 CPM for a thousand general impressions collapsed.

Not because brand impressions weren’t valuable — the research consistently supports brand awareness — but because the organizational incentive structure of most B2B marketing teams rewards pipeline metrics, not brand metrics. Demand generation managers have quotas. Brand managers have budgets. Quota-holders win budget fights.

The CPM didn’t just fall in value. It lost its job description.

Then the model ate itself, in two ways that nobody at the time saw coming clearly enough.

1. The race to the bottom

Once guaranteed lead generation became standard, low-cost providers flooded the market. Publishers who had carefully built niche, qualified audiences suddenly found themselves competing against data brokers and content farms offering leads at $15 or $20 CPL, a fraction of what quality audience-based programs cost to produce. Vendors promising guaranteed lead volumes at fixed CPLs were often selling someone else’s database with a light content wrapper around it. Advertisers, under relentless pressure to hit MQL targets at lower cost, chased the cheaper option. Quality collapsed along with price. Some marketers eventually dropped syndication efforts entirely because of poor lead quality, but not before the race to the bottom had compressed CPLs across the entire category, including for publishers who were actually delivering quality.

2. The List Burnout Problem

The second structural damage was quieter but more corrosive. Publishers who used their own first-party subscriber lists to fuel lead gen programs, the logical thing to do since their owned audiences were their most valuable asset, discovered they were slowly burning those lists down. A subscriber who downloads three white papers on behalf of different advertisers in a quarter stops opening emails. A professional who gets called by four vendors within 48 hours of filling out a form learns to use a junk email address. The audience that had taken decades to build and years to earn began eroding under the very programs designed to monetize it. The core asset was being consumed to generate short-term revenue. Publishers largely didn’t see it coming until open rates had already fallen and unsubscribe rates had quietly climbed.

A smarter tier eventually emerged. High-quality content syndication providers built what the commodity-players couldn’t replicate: genuine first-party data marketplaces. Quality audiences. Verified professional identities. Permission-based, GDPR-compliant data. Behavioral intent signals built from real engagement rather than form-fill arbitrage. Companies operating at this level could command $80–$150 CPL’s — sometimes more — because their leads actually converted. They proved that quality and compliance at scale were achievable. But they also revealed an uncomfortable truth: most publishers couldn’t get there. The audience depth wasn’t there. The data infrastructure wasn’t there. The discipline to protect list health over short-term revenue hadn’t been there when it mattered most. The premium tier thrived. The commodity tier continued racing downward. And the publishers caught in the middle — too principled to go fully commodity, too under-resourced to build a genuine data marketplace — found themselves with the worst of both worlds.

The tragedy of the lead gen era is that it gave B2B publishers a genuine revenue lifeline at exactly the moment digital CPMs were collapsing. Through a combination of market pressure and short-term thinking, the model undermined the very audience equity it was supposed to monetize.

The Three Paths Forward

Legacy B2B digital publishers have three viable strategic options. Each has merit under specific conditions. None is without significant risk.

Option 1: Ride the Declines

Accept the trajectory and optimize for maximum cash extraction from existing operations. Reduce editorial headcount, consolidate brands, automate content production with AI, and manage to margin rather than growth.

This is the media industry’s version of a harvest strategy. Extract value while the asset still has it, and don’t invest unncessary capital into a declining business.

• The case for it: No capital required. Predictable near-term cash flow. AI dramatically reduces content production costs. Clean operational simplicity.

• The case against it: Editorial quality erosion accelerates the audience decline you’re managing. Destroys asset value for any future exit. PE investors rarely tolerate it once they understand what it is. And it’s only viable with low debt ratios which rules out many of the roll-up portfolios carrying mezzanine financing.

Option 2: Double Down on Events

Shift the center of gravity from digital advertising to in-person events, which command dramatically higher effective CPMs. Event sponsorships, exhibition fees, and curated access to peer communities can yield $50–$200+ in per-attendee-equivalent revenue: 10x–30x programmatic advertising rates.

The argument is sound: events leverage the one thing digital can’t replicate. Physical presence, serendipitous conversation, curated peer access. The community is the product.

• The case for it: Premium sponsor economics vs. digital. Defensible community moat. AI can’t yet replicate in-room value. High attendee NPS when executed well.

• The case against it: High fixed costs and operational complexity make the math unforgiving. The events market is bifurcating sharply. Only the top one or two franchises in each vertical survive at scale, while second-tier events collapse under their own cost structure. Vulnerable to economic downturns. It doesn't scale like digital inventory.

Option 3: Build the Data Layer

Transform the first-party audience data — accumulated through decades of editorial relationships — into a proprietary intelligence and targeting product.

Build a unified data layer across the brand portfolio. Package it for ABM and intent-targeting buyers. Sell audience access at $25–$75 CPM rather than $5–$7. Layer in research, intelligence reports, and predictive intent signals as standalone revenue streams. Become the vertical’s data infrastructure, not just its content platform.

• The case for it: CPMs 5x–10x the open exchange rate. Genuinely defensible if the data is deep and structured. Aligns with where B2B marketing budgets are actually moving. Has SaaS-like revenue potential at scale.

• The case against it: Requires significant technology investment most publishers haven't made. First-party data is fragmented across brands — nowhere near a unified platform. The sales force must be retrained entirely. Selling data products is structurally different from selling media placements. And it takes 2–4 years to monetize properly. The interim period is brutal.

Where Legacy B2B Media Ultimately Lands: A Prediction

Here’s what I think actually happens over the next five years, as agentic AI accelerates every trend described above.

The survivors become Vertical Intelligence Platforms. A small number of legacy B2B media brands — those who move aggressively to own first-party data, build community infrastructure, and position as authoritative vertical intelligence platforms — will survive and potentially thrive. Agentic AI needs trusted, structured, domain-specific data to operate effectively. Publishers who become “that source”, rather than competing with AI for search traffic, find a new and genuinely valuable role: the credentialed data provider behind the answer engine. Think less “publisher,” more Bloomberg Terminal for niche verticals.

The majority of roll-up portfolio brands face an existential reckoning. Those running on thin editorial teams, generic programmatic inventory, and me-too digital advertising products face an existential reckoning by 2028. Agentic AI disintermediates their search-driven traffic. CPMs continue eroding as supply expands and buyer sophistication increases. Without a genuine audience relationship or differentiated data asset, they are advertising inventory businesses in a world of infinite advertising inventory. The PE math stops working. Assets get written down or sold for parts.

The most underestimated asset is community. Not content. The deepest moat in legacy B2B media isn’t the content. It’s the community. Decades of industry relationships, trusted editorial voices, professional peer communities. Conferences. Awards. Roundtables. This social infrastructure is genuinely hard for AI to replicate. Publishers who recognize they are fundamentally “community businesses” and price, build, and operate accordingly, may find the most durable path forward.

Because here’s the thing: the content was always just a reason to gather. In a post-AI world, the gathering itself becomes the product.

The Question That Cuts to the Center

The B2B media industry is not dying. But the B2B media business model, built on CPM advertising against editorial traffic, certainly is.

What replaces it will look more like a data business, a community business, or a marketing services business than anything a traditional publisher would recognize.

The publishers who survive will be those who understand they were never really in the media business at all.

They were in the professional audience business.

That business still has a future.

The question is whether they move fast enough to find it.

Subscribe now

The views expressed in Uphoff on Media are entirely my own. They don’t represent the opinions of any company I’ve led, any board I’ve sat on, or any investor who’s had the pleasure of debating strategy with me over the years. If something I write here sounds brilliant, I’ll take full credit. If it turns out to be wrong, I was clearly misquoted by myself.

By Tony Uphoff | Uphoff on Media

Open LinkedIn on any given morning and you’ll find a fresh wave of “AI is coming for your job” posts, usually written by someone selling an AI course or a consulting engagement. The comment sections fill with a predictable mix of anxiety and bravado. The anxious ones worry they’ll be replaced by a chatbot. The bravado crowd insists AI is “just a tool” and nothing will really change. Both camps are missing the point.

AI isn’t coming for all the jobs. It’s coming for certain kinds of work — and the distinction between the work that survives and the work that doesn’t is more nuanced than most people want to admit.

As a five-time CEO, I’ve led companies through multiple major market transitions over 35 years. I’ve watched newsrooms go digital, print advertising collapse, and entire business models rebuilt from scratch. Every single time, the initial panic about job losses was both understandable and overstated. And every single time, the actual reshaping of roles and responsibilities was more profound, and more interesting, than the headline layoff numbers suggested.

This time is no different. And also, this time is different. Let me explain.

First, Let’s Acknowledge What’s Actually Happening

The numbers are real and they’re uncomfortable. In 2025 alone, over a million job cuts were announced across U.S. industries. The highest level since the pandemic. Tech companies cut roughly 127,000 workers. The advertising and marketing industries saw meaningful contraction, with agency head-counts falling an estimated 8% and job openings in advertising and marketing declining more than 7% between 2022 and 2025. January 2026 brought the worst month for layoff announcements since the Great Recession.

These are real people, real families, and real careers being disrupted. I’ve been on both sides of that table. I’ve made the decision to restructure teams, and I’ve watched talented people walk out the door knowing their skills didn’t fit the new model. There is nothing abstract about this when it’s happening to you or someone you care about.

In a post published earlier this week, I introduced the First Mile / Last Mile framework ; the idea that humans own the early work of reading signals, decoding buyer behavior, and translating insight into strategy, while AI owns execution and scale. If that framework holds, and I believe it does, it raises an immediate and uncomfortable question for every B2B marketing leader: does your current org chart reflect that division of labor? For most teams, the honest answer is no. The structure was built for a different era. One where humans executed and scale was the constraint. That era is over.

So let me be direct: I’m not going to sugarcoat what’s coming with false reassurance. But I am going to push back on the hysteria, because the hysteria is just as dangerous as the complacency.

Subscribe now

The Overcorrection Toward Efficiency

Here’s what I think the “AI is destroying all the jobs” narrative misses: a significant portion of what we’re seeing right now isn’t AI-driven disruption. It’s a market correction masquerading as a technology revolution.

We just came through the most extraordinary period of cheap money and growth-at-all-costs thinking in modern business history. Between 2020 and 2022, B2B marketing budgets ballooned to nearly 10% of company revenue. Companies hired aggressively, stacking teams with specialists, coordinators, and managers for every conceivable function. The logic was simple: money was essentially free, growth was the only metric that mattered, and headcount was a proxy for ambition.

Then interest rates rose, venture capital dried up, and Wall Street rediscovered the concept of profitability. Suddenly, every CEO and CFO in America started looking at their org charts through a very different lens. AI became the convenient justification for cuts that were actually driven by financial discipline. “We’re leveraging AI to improve efficiency” sounds a lot better in an earnings call than “we overhired during the bubble and now we’re correcting.”

That’s not to say AI isn’t a genuine driver of change, it absolutely is. But we need to separate the cyclical correction from the structural transformation if we want to understand what’s really happening and, more importantly, what’s coming next.

Because here’s the part that should actually make you optimistic: the efficiency phase is temporary. The growth phase is what follows. And when AI shifts from being a cost-cutting story to a revenue-growth story, the demand for talent doesn’t disappear, it transforms.

The Hollowing of the Middle

So what does the structural transformation actually look like? In my experience leading marketing and media organizations through transitions, the pattern is remarkably consistent: the middle gets hollowed out.

Every B2B marketing organization I’ve run or worked with has a similar architecture. At the top, you have strategic thinkers; the people setting direction, defining positioning, making bets on where the market is going. Closest to the machine, you have the technical builders, the people who design systems, architect data flows, and make the machinery actually work. And in the middle, you have a large layer of people whose primary job is to translate strategy into execution: coordinating campaigns across platforms, scheduling content, pulling reports, reconciling data between systems, managing workflows between teams, and generally keeping the trains running on time.

That middle layer is exactly what agentic AI replaces. Not because those people aren’t smart or capable. Most of them are. But because the work itself is fundamentally about sequencing tasks, moving information between systems, and managing processes. That’s precisely what AI agents are built to do, and they do it faster, cheaper, and around the clock.

This is not a future prediction. It’s happening now.

The Roles That Decline

Let me be specific, because generalities don’t help anyone plan their career.

• Campaign operations managers who manually build and schedule campaigns across email, social, paid media, and content platforms. When an AI agent can orchestrate a multi-channel campaign from brief to execution, adjusting targeting, timing, and creative in real time based on performance signals, the human coordinator becomes redundant. Not the strategy behind the campaign. The coordination of it.

• Content production specialists focused on volume. The blog writers churning out three SEO-optimized posts per week, the social media schedulers maintaining editorial calendars, the copywriters producing variations of the same messaging for different channels. AI generates competent content at scale. “Competent at scale” is exactly the value proposition these roles were hired to deliver. The math doesn’t work in their favor.

• Marketing ops analysts who spend their days pulling reports from multiple platforms, reconciling data across systems, and building dashboards. This is precisely the kind of structured, repetitive, cross-system work that AI handles natively. The insight that comes from analyzing the data is valuable. The manual labor of assembling it is not.

• Media planners who allocate budgets across channels based on historical performance and industry benchmarks. AI-driven media optimization already outperforms human planners on quantitative allocation decisions. The strategic judgment about which channels align with brand positioning still matters. The spreadsheet work of dividing dollars across line items does not.

• Entry-level SDR and BDR roles built around scripted outbound activity. AI handles the research, personalization, sequencing, and initial outreach with more precision and persistence than a junior sales rep working a call list. The human handoff point moves further down the funnel, from initial outreach to qualified conversation.

If you’re in one of these roles, I’m not telling you to panic. I am telling you to start planning your evolution now, and I’ll come back to that.

The Roles That Rise

Here’s the part that gets less attention but matters more: the roles that become dramatically more valuable in an AI-driven marketing organization.

Brand strategists and creative directors. This might seem counterintuitive. Isn’t AI generating creative now? Yes, and that’s exactly the point. When AI floods every channel with competent content and creative, the premium on distinctive brand voice, positioning, and creative direction doesn’t just hold, it skyrockets. Competence becomes commodity. Distinction becomes the differentiator. The people who can define what a brand sounds like, looks like, and stands for, and who can ensure that AI-generated content actually reflects that distinctiveness, become the most valuable people in the building.

Data and AI architects. The “garbage in, garbage out” problem doesn’t go away when you add AI. It gets amplified exponentially. An AI system trained on bad data or poorly structured information doesn’t just produce bad results; it confidently produces bad results at scale. The people who design data architectures, ensure data quality, select and train models, and build the infrastructure that AI operates on become the foundation of everything else. This is one of the most in-demand and underdeveloped skill sets in B2B marketing today.

Revenue strategists. People who can can think across the full buyer journey, from interest-to-invoice. Not just the marketing funnel, not just the sales pipeline, but the complete arc from initial signal to closed deal to customer expansion. The organizational walls between marketing and sales have always been artificial, but they persisted because the systems were separate and the data didn’t flow. AI removes that excuse. The people who can think holistically about revenue generation across the entire cycle become indispensable.

Customer insight specialists. AI generates vast amounts of behavioral data and intent signals. But data without interpretation is just noise. The people who can read faint signals, decode the why behind shifting buyer behavior, and translate insight into executable strategy, they become the bridge between machine capability and human judgment.

AI governance and ethics roles. Someone has to ensure the autonomous systems are operating within brand guidelines, legal requirements, regulatory frameworks, and ethical guardrails. When an AI agent is sending thousands of personalized outreach messages on your behalf, who’s making sure they’re accurate, compliant, and consistent with your brand values? This role barely exists today. Within two years, it will be a requirement for any serious B2B marketing organization.

The Rise of the Chief Growth Officer

All of this leads to what I believe is the most significant structural change coming to the B2B marketing org chart: The convergence of marketing and sales leadership into a unified growth mandate.

Here’s the data that tells the story. According to Spencer Stuart’s research, average CMO tenure at Fortune 500 companies sits at roughly 4.1 years. One of the shortest tenures in the C-suite, well below the CEO average of 6.7 years. Forrester’s data shows it declining further, from 4.1 years in 2024 to 3.9 years in 2025. Meanwhile, the percentage of Fortune 500 companies with a marketing executive who reports directly to the CEO or sits on the leadership team dropped from 63% to 58% in just one year.

And here’s the detail that tells you where this is heading: only 40% of Fortune 500 marketing leaders actually hold the title of Chief Marketing Officer. The rest carry titles like Chief Growth Officer, Chief Revenue Officer, Chief Customer Officer, or hybrid combinations that reflect a broader mandate than traditional marketing alone. More than a quarter of B2B firms introduced new senior marketing roles in just the past year: including Chief Growth Officer at 4% and Chief Revenue Officer at nearly 4%.

This isn’t a cosmetic rebranding exercise. It reflects a fundamental organizational recognition that the separation of marketing and sales is an artifact of a pre-AI era. When agentic AI can orchestrate the entire journey from intent signal to qualified opportunity to closed deal, adjusting messaging, channel mix, and engagement strategy in real time, the organizational boundary between “marketing” and “sales” stops making operational sense.

The Chief Growth Officer isn’t a rebranding of the CMO. It’s a fundamentally different mandate. The CGO owns the full revenue cycle: demand creation, pipeline development, sales enablement, customer expansion. They operate with shared data, shared systems, and shared accountability for outcomes. They don’t hand off leads from marketing to sales. They orchestrate a continuous revenue process that AI manages across the entire buyer journey.

I know this will provoke a reaction from CMOs reading this. Good. Here’s my honest take: the standalone CMO role, as currently defined in most B2B companies, is heading toward one of two fates. Either it gets elevated into a CGO-type growth mandate — absorbing sales leadership and owning the full revenue cycle — or it gets reduced to a communications and brand function that reports into the CGO or CEO. The middle ground, where the CMO owns “marketing” as a separate function with separate metrics and separate accountability, is the position that AI makes untenable.

If you’re a CMO today, the question isn’t whether this convergence is coming. The question is whether you’re going to lead it or be reorganized by it.

The Hysteria Is Overhyped. The Transformation Is Not.

Let me be clear about something: I think a lot of the “AI will replace everyone” narrative is clickbait. Not because AI isn’t transformative. It is. But because the people pushing the most extreme versions of this story are usually selling something, and because history tells us that technology transitions create as many roles as they eliminate. The roles just look different.

We saw this in media. When digital disrupted print publishing, the “print is dead, all journalists will lose their jobs” narrative was everywhere. What actually happened was more nuanced. Yes, thousands of print-centric roles disappeared. But entirely new categories of work emerged: digital content strategy, audience development, social media management, data journalism, programmatic advertising operations. The total number of people employed in media changed less dramatically than the composition of those roles.

The same pattern is playing out in B2B marketing. The total demand for human talent in marketing won’t collapse. But the kind of talent that’s valued, and the way that talent is organized, will shift dramatically.

And here’s the part that should make you genuinely optimistic: the current wave of efficiency-driven cuts is a phase, not a destination. Wall Street is rewarding cost reduction right now because we’re in a correction from the growth-at-all-costs era. But the companies that only use AI to cut costs will eventually lose to the companies that use AI to grow revenue, improve margins, and create competitive advantages. When that shift happens — and it’s already beginning — the demand for people who can architect AI-driven growth strategies, interpret complex data, build distinctive brands, and lead cross-functional revenue teams will outstrip supply.

The winners won’t be the companies with the fewest people. They’ll be the companies with the right people doing the right work.

What This Means for You

If you’re reading this and thinking about your own career, here’s my practical advice.

Invest in self-education now. The window to get ahead of this transition is open, but it won’t stay open forever. Understand how agentic AI works. Not at the technical level, but at the operational level. How does it change campaign planning? What does it do to the buyer journey? How will it affect your specific function? The people who walk into their next performance review with a thoughtful perspective on how AI will reshape their role, and a plan for how they’ll evolve with it, will be the ones who survive and thrive. The ones who wait for their company to train them will be waiting too long.

Move toward the edges. If you’re in that middle layer I described, the coordination and execution zone, start building skills that push you toward either the strategic end or the technical end. Learn to think about brand positioning and marketing strategy, not just campaign execution. Or lean into data architecture, AI systems, and the technical infrastructure that powers modern marketing. The middle is where the compression happens. The edges are where the value accrues.

Think about revenue, not marketing. The fastest way to future-proof your career is to start thinking and talking in terms of revenue impact, not marketing metrics. Pipeline, conversion, customer lifetime value, expansion revenue. These are the languages that will define the next generation of marketing leadership. If your vocabulary is still dominated by impressions, clicks, MQLs, and engagement rates, you’re speaking a language that’s losing its currency.

I’ll be writing more about career management in the age of AI in future posts. It’s a topic that deserves deeper treatment than I can give it here. For now, the most important thing I can tell you is this: The people who treat this transition as something happening to them will struggle. The people who treat it as something they can actively navigate, with curiosity, self-education, and a willingness to evolve, will find that the AI era creates more opportunity than it destroys.

It just won’t create opportunity for the same work we’ve been doing.

Subscribe now

The views expressed in Uphoff on Media are entirely my own. They don’t represent the opinions of any company I’ve led, any board I’ve sat on, or any investor who’s had the pleasure of debating strategy with me over the years. If something I write here sounds brilliant, I’ll take full credit. If it turns out to be wrong, I was clearly misquoted by myself.

Uphoff on Media · Frameworks Series

There is a conversation happening right now in every B2B marketing organization worth its budget. It goes something like this: How much of what we do can be done by AI agents? What should humans actually own? And, the question nobody wants to say out loud, are we building toward a future where half this team is optional?

I've spent 35 years at the intersection of B2B media, marketing, and technology. I've watched the internet rewire publishing, SaaS rewire software, cloud rewire the enterprise stack, and mobile rewire everything else. And I’ll tell you what I tell every executive who asks me some version of this: you’re asking the wrong question.

The right question isn’t how much AI replaces. It’s where are humans irreplaceable. And the answer, I’d argue, is hiding in a concept borrowed from logistics.

In the agentic AI era, human judgment owns the First Mile and the Last Mile. The middle miles belong to the machines. Get this wrong, and you’ll either under-invest in AI or abdicate judgment you can’t afford to lose.

The Coach Who Never Played the Game

Before I lay out the framework, let me give you a mental model that’s been useful in my own thinking.

Consider the professional coach. Not the player, the coach. Some of the greatest coaches in sports history were mediocre players, or never played professionally at all. Bill Walsh, architect of the 49ers dynasty, was a career backup. Phil Jackson was a journeyman forward who averaged eight points a game. What they had wasn’t athletic superiority. It was something different: the ability to see the whole system, prepare the team, call the right play at the right moment, and make the adjustments that winning requires.

That’s the job of B2B marketing leaders in an agentic AI world. You don’t have to run every route. You have to know which routes to run, why, against which defense, and what to change at halftime when the first-quarter data tells you something unexpected.

The AI agents? They’re the players. Fast, tireless, scalable. But they run the plays you call. And they can’t read the faint signals that tell you the game is shifting before the scoreboard shows it.

The Framework: First Mile, Middle Miles, Last Mile

Let me make this concrete. In logistics, the First Mile is getting the product from the manufacturer to the distribution network. The Last Mile is getting it from the distribution center to the customer’s door. Both are notoriously hard, expensive, and human-intensive. The middle — the highway miles — is where efficiency and automation dominate.

The same structure maps cleanly onto agentic AI-powered B2B marketing.

The First Mile: Human Judgment at the Point of Intent

The First Mile is everything that happens before you deploy an AI agent on a task. This is where human judgment is not just useful, it’s the entire game.

First Mile work includes:

• The What and the Why. What are we actually trying to accomplish, and why does it matter to this customer, in this market, at this moment? AI agents are extraordinarily good at executing direction. They are not good at questioning whether the direction is right. That’s your job.

• Brand Truth and Market Truth. Every B2B organization has a set of things that are true about their brand; how it’s perceived, what it stands for, where it’s earned credibility and where it hasn’t. Equally important is Market Truth: what customers actually believe, what competitors are doing, where the category is heading. These truths aren’t in your CRM. They live in human judgment built from years of experience. An AI agent handed bad Brand Truth or outdated Market Truth will execute beautifully in the wrong direction.

• Faint Signal Recognition. This one is wildly under-appreciated. Some of the most important information in any market exists before it’s legible to large language models, before it’s in enough published sources to influence training data, before it shows up in search trends, before your competitors see it. Customer sentiment shifts. Emerging vocabulary in niche communities. An account team’s gut read that something has changed in a key relationship. These faint signals are a human intelligence advantage that evaporates the moment you remove experienced judgment from the First Mile.

• The How. Not the tactical execution, that’s for agents. The strategic how: Which channels carry the message? What sequence? What’s the right tone for this audience segment versus that one? What does this campaign need to feel like? These are aesthetic and strategic decisions that require the kind of taste that comes from having been wrong a few times and learning from it.

Feed an AI agent bad First Mile inputs and you get a highly efficient machine producing precisely the wrong outcomes. Garbage in, garbage out, just faster than ever before.

The Middle Miles: Where Agents Earn Their Keep

Once the First Mile is right, once intent is clear, truth is established, signals are read, and direction is set, the Middle Miles are where agentic AI delivers its extraordinary value proposition.

Research and synthesis at scale. Content generation, variation, and personalization across segments. Campaign execution across channels. A/B testing and optimization loops. Data aggregation and pattern recognition across datasets too large for any human team to process. Scheduling, sequencing, and deployment. Reporting and performance monitoring.

None of this requires human judgment in real time. All of it benefits from being done faster, at greater scale, and with more consistency than a human team can deliver. This is the legitimate promise of agentic AI in B2B marketing. And it’s a substantial one. I’m not dismissing it. I’m contextualizing it.

The mistake most organizations are making right now is trying to automate their way into the First Mile. They’re deploying agents earlier and earlier in the process, treating strategic intent as something that can be templated, and discovering, sometimes painfully, that speed without direction is just an accelerated path to the wrong place.

Subscribe now

The Last Mile: Human Judgment at the Point of Truth

The Last Mile is where the work comes back to humans, and where most frameworks for AI integration stop short.

Last Mile human judgment includes:

• Taste and Curation. Your AI agent can generate fifty variations of a thought leadership piece. A human has to decide which one is actually good, which one sounds like your brand, earns the reader’s trust, and doesn’t have that slightly unnatural quality that sophisticated B2B buyers have already learned to detect. Taste is a human function. Full stop.

• Accuracy and Brand Integrity. LLMs hallucinate. They confidently assert things that are partially right, subtly wrong, or outdated. In B2B marketing, where you’re talking to buyers who know as much-if not more- about your category than you do, a factual error or a claim that doesn’t hold up isn’t just embarrassing, it’s brand-damaging. Human review at the Last Mile isn’t optional. It’s the quality control layer that makes everything else trustworthy.

• Customer Truth. AI agents can analyze customer data. They cannot replicate the judgment of a senior marketer or account executive who has spent years in genuine relationship with a customer. Who knows that a particular buyer distrusts a certain kind of claim, or that an account is quietly evaluating alternatives, or that a relationship that looks fine on paper is actually fragile. That knowledge lives in humans, and it needs to apply at the Last Mile before anything goes out the door.

• The Judgment Call. Sometimes the data points one way and experience points another. Sometimes the AI-generated output is technically correct but feels wrong. Sometimes the campaign as designed is solid but the timing is off. These are judgment calls. They require a human who has enough context, experience, and accountability to make them, and own them.

The Last Mile isn’t where you check the AI’s work. It’s where you apply the judgment that no model can replicate, and where the difference between good marketing and great marketing actually lives.

What This Means in Practice

If you’re leading a B2B marketing organization right now, here’s how I’d suggest applying this framework:

1. Audit your current AI deployment against the framework. Where are you using agents? Are they operating in the Middle Miles, where they belong, or have they crept into First Mile strategic decisions or replaced Last Mile human review? The creep is subtle and usually well-intentioned. It’s also where the risk concentrates.

2. Invest in First Mile capability, not just AI tools. The bottleneck in most agentic AI deployments isn’t the technology. It’s the quality of human inputs going in. Senior judgment, market knowledge, Brand Truth documentation, customer intelligence, these are the inputs that determine whether your AI investment delivers or disappoints.

3. Build a Last Mile review process that’s actually resourced. Not a rubber stamp. Not a five-minute skim before publication. A genuine human review function staffed by people with the taste, knowledge, and authority to say no. This is where you protect your brand in an era when AI content is everywhere and differentiation increasingly comes from quality and trust.

4. Name your faint signals. Make it someone’s explicit job to surface and document the early-stage customer and market intelligence that hasn’t yet made it into structured data. The signals that won’t show up in your analytics dashboard for another six months but that experienced humans in your organization can already sense. That intelligence is a competitive asset. Treat it like one.

The Human Advantage, Precisely Located

The debate about AI and human work has been framed, badly, as a binary one. Replace or don’t replace. Automate or don’t automate. That framing is useless because it’s imprecise. The reality is, we now live in a world of Human + Digital labor. And we’re not going back.

The more useful question, the one that actually helps you make decisions, is: at which points in this process is human judgment genuinely irreplaceable, and at which points is it just an expensive habit?

The First Mile and Last Mile framework gives you a principled answer. Humans own intent, truth, taste, faint signals, and the final judgment call. Agents own the execution in between. That’s not a concession to AI. It’s a clear-eyed map of where the human advantage is actually located.

AI can generate fifty options. It cannot tell you which one is right. That distinction, between output and judgment, is the whole game. Your job isn't to prompt better. It's to decide better. That's not a skill the model has. It's the one you bring.

Subscribe now

The views expressed in Uphoff on Media are entirely my own. They don’t represent the opinions of any company I’ve led, any board I’ve sat on, or any investor who’s had the pleasure of debating strategy with me over the years. If something I write here sounds brilliant, I’ll take full credit. If it turns out to be wrong, I was clearly misquoted by myself.

My standard sign-off in texts and emails has always been “Best, Tony.” Simple. Professional. Warm enough.

During the years I was running UBM TechWeb: a global B2B media, events, and data business that grew to over $300 million in revenue, that sign-off had a problem. I wasn’t always “Best, Tony.” Thanks to iPhone autocorrect, occasionally I was “Beast, Tiny.”

The irony is hard to miss. At 6’5” and 200 pounds, “Tiny” is not a word anyone would use to describe me. And I’ve spent my career building a leadership style and reputation around decisiveness with empathy, clarity with care. Not exactly “Beast” territory.

But here’s what I learned: autocorrect doesn’t lie. It just exposes what you’re actually doing.

The Pressure Cooker

The years of the Beast/Tiny phenomenon were not ordinary years.

The smartphone was rewriting how business got done in real time. At UBM TechWeb, we were simultaneously managing the decline of print, scaling digital, acquiring and globalizing events, including operations in Mumbai, Beijing, and Shanghai. And then the global financial crisis hit, compressing timelines, rerouting capital, and forcing us to rewire the planes while in flight.

I was texting constantly. To direct reports, to global teams, to board members, to partners. From airports, hotel rooms, back-seat rides to meetings. Direction, questions, answers, decisions, all of it compressed into a thumb-typed stream of consciousness. My texts were moving at the speed of the market.

And the market was not waiting for me to proof my sign-off.

The Loneliest Job You’ll Never Fully Explain

Here’s what doesn’t get talked about enough: running a company at that pace, in that environment, is genuinely isolating.

The data on this is striking. Research shows that half of CEOs and senior leaders report experiencing loneliness and isolation, with 61 percent believing it negatively impacts their performance. Senior leaders are twice as likely to report feelings of isolation compared to employees in non-leadership positions.

I lived that statistic without knowing it existed.

The isolation isn’t social, you’re surrounded by people. It’s contextual. As CEO, you’re the one everyone looks to for context. Which means fewer and fewer people can give it to you. Information flows up filtered. Perspective flows up managed. The loneliness experienced by CEOs stems not from a lack of social connections but from the heavy burden of leadership and decision-making; especially during crises, when they look to their board, senior executives, or operational managers and find that even those people are watching what they say.

The faster the market moved, the more I leaned on digital communication to stay connected. The more I leaned on digital, the more isolated my actual thinking became. I was generating output — decisions, direction, responses — without adequate input. A machine running hot with no coolant.

Autocorrect caught what I couldn’t. I was moving too fast. And “Beast, Tiny” was the tell.

Subscribe now

The Framework: Three Places the Gap Shows Up

What the Beast/Tiny mistake revealed was a gap, between intent and output. That gap isn’t unique to texts. Under sustained pressure, it shows up across three domains every leader should monitor.

1. Communication: What You Meant to Say vs. What Landed

Speed collapses nuance. Autocorrect made it visible; usually it isn’t.

Most leaders believe they’re communicating clearly under pressure. Research suggests otherwise. Studies on managerial communication consistently find that during high-stress periods, message clarity degrades while message volume increases. A particularly dangerous combination in a global organization where context is already challenged across time zones and cultures.

The discipline: Before sending anything consequential, ask one question: “What will this person do with this?” Not what you meant. What they’ll do. That five-second check is the equivalent of proofreading your texts. It costs nothing. The cost of skipping it compounds.

2. Decisions: What You Intended to Signal vs. What the Organization Heard

Fast-moving leaders often experience themselves as decisive. Their organizations often experience them as reactive, and the difference is enormous.

A decisive decision has visible reasoning. A reactive decision has visible urgency. When you’re operating in crisis mode, urgency can masquerade as clarity. People execute, but they execute on what they inferred, not what you intended.

Former U.S. Surgeon General Vivek Murthy noted that loneliness “reduces task performance, limits creativity, and impairs other aspects of executive function such as reasoning and decision making.” When isolation compounds velocity, the decisions that feel sharpest may actually be the least examined.

The discipline: After significant decisions, do a 24-hour check. Not to reverse, but to listen for what the organization is doing with what you said. If the actions don’t match the intent, the gap is yours to close.

3. Presence: Who You Meant to Be in the Room vs. Who Actually Showed Up

This is the hardest one.

At speed, leadership presence defaults. You stop showing up intentionally and start showing up habitually, running on pattern, not purpose. The “Beast” isn’t aggressive. The Beast is the unexamined version of you, operating on autopilot while you manage a hundred other inputs.

Fewer peers, high external expectations, and rare honest feedback make it difficult to openly discuss one’s limitations. Most CEOs don’t get told when their presence has shifted. They find out later; in an exit interview, a board conversation, a candid moment from a trusted advisor, that people had been reading a version of them they hadn’t intended to project.

The discipline: Build at least one feedback loop that is structurally honest, a peer, an executive coach, a trusted board member, someone whose job is not to manage your reaction. The Beast/Tiny moment happened because someone on my team felt comfortable enough to laugh about it with me. That comfort didn’t happen by accident.

What “Beast, Tiny” Actually Fixed

When people caught the autocorrect mistake, they loved it. Initially I was embarrassed. Then I realized what it was doing: it created permission to be human.

“Beast, Tiny” broke the artifice of the relentlessly-in-command CEO. It reminded my team — and honestly, reminded me — that the person sending these texts was a human being making judgment calls under real pressure, not an oracle issuing pronouncements from on high.

The forcing function wasn’t just proofreading my sign-offs. It was learning to pause before I communicated, before I decided, before I walked into a room. Not slower, more deliberate. The difference matters.

The market was moving at autocorrect speed. Leadership couldn’t.

The Test You Can Apply Today

I now think of the Beast/Tiny Test as a simple diagnostic leaders can run in three moments:

• Before you send: Does this say what I mean, or what I typed?

• After your decision: Is the organization moving the way I intended, or the way they interpreted?

• Before you enter the room: Am I showing up on purpose, or on pattern?

Three questions. Seconds each. The cost of skipping them is what autocorrect was trying to tell me.

Best,

Tony

Subscribe now

The views expressed in Uphoff on Media are entirely my own. They don’t represent the opinions of any company I’ve led, any board I’ve sat on, or any investor who’s had the pleasure of debating strategy with me over the years. If something I write here sounds brilliant, I’ll take full credit. If it turns out to be wrong, I was clearly misquoted by myself.