AI Has Changed the Way B2B Buyers Buy. Just Not the Way You Think

Everything I've argued in recent posts came from the operator's chair, reading the market from the supply side: that B2B media's future belongs to companies that go deep instead of broad, and that events are the one format agentic AI can't disintermediate.

Last week I got the buyer’s side of the story, and it didn’t come from a hunch or a hallway conversation at a conference. It came from Outsell, in their new B2B Buyers Study: based on a survey of 226 B2B buyers who control real spend.

I read it looking for reasons to push back on my own arguments. There weren't many to find: the buyers confirmed, with hard numbers, two concepts I've been thinking and writing about from the outside looking in. But there was something neither of my previous posts covered, because I didn't have this data yet: a problem sitting inside that confirmation that ought to worry anyone running a company that buys, not just the ones trying to sell to it.

Where This Data Came From

I've been an Outsell customer twice: as CEO of UBM TechWeb, and again at ThomasNet.com, the company I cited recently as one of the clearest examples of the vertical intelligence model in practice. I paid for Outsell's research in both jobs because it was buyer-centered work built for operators. This time is different: Outsell sent me this study unsolicited and at no cost and encouraged me to share my own take.

This study was done in partnership with independent researcher Rebecca Kitchens, a 23-year TechTarget veteran who ran the business through its combination with Informa. It's Outsell doing what Outsell does: the research and advisory firm for B2B information, data, and media companies, delivering work built for the people running the business. It's also rare. I haven't seen another study that goes directly to B2B buyers and asks what's actually happening when AI enters the purchase process, instead of inferring it from traffic data. If it exists, I haven't found it, and I've been looking. The industry needs more of this kind of data and analysis.

The Depth Argument, Confirmed by the People It’s Supposed to Serve

In “The Vertical Intelligence Company”, I argued that the businesses winning the machine age are going deeper into a specific professional domains instead of broader across many, and that buyers were rewarding depth with trust they wouldn’t extend to a generalist. Outsell’s data says exactly that, from the buyer’s side of the transaction.

This one surprised me at first: third-party source reliance rising at the same moment AI adoption is felt counter-intuitive. Reliance on third-party sources is rising, not falling, since buyers started using AI. Peer review sites and peer communities are both up roughly 49%. Analyst reports are up 46%. Trade publications are up 42%. The only category moving the other direction is vendor-produced content, down across the board and down hardest, 14%, among the buyers who'd normally be closest to a sales conversation. When AI can synthesize a vendor's own positioning in seconds, going straight to the most biased source in the room stops making sense.

Rising reliance is the headline. It’s not the finding that matters most.

The finding that matters most is where in the funnel that reliance now sits, and it’s the one I’d want every media and event operator to think about.

Buyers aren’t coming to third-party sources to get oriented anymore. AI does that job now. They’re coming later, after AI has already narrowed the field, to check what they were told and find what AI left out. Same sources, more reliance, a completely different job in the funnel.

That reframes the traffic story most operators are telling themselves. A decline at the top of the funnel isn’t your audience leaving. It’s AI absorbing the orientation work your homepage, your explainers, and your top-of-funnel content used to do. The audience that’s left has moved downstream: later in the process, harder to catch with reach-driven tactics, and paying closer attention when they do show up, because they’re there to verify a shortlist, not to browse.

And zero respondents, out of 226, said they use AI exclusively in vendor selection. Not one. AI didn’t replace the research ecosystem. It became the front door to it. Buyers are walking through that door faster, then heading straight for the same trusted rooms they always used, just being pickier about which ones once they get there, and arriving later in the conversation than they used to.

The Events Argument, Confirmed With a Number I Didn’t Expect

In “The B2B Event Market Is Splitting in Two” , I argued events are the one revenue line agentic AI can’t touch, because it can synthesize a panel but it can’t replicate a room. I expected the buyer data to support that argument. I didn’t expect it to be the strongest number in the entire study.

Events had the single largest increase in reliance of any source category Outsell measured: 53%, ahead of peer communities, analyst reports, and trade publications. Buyers aren’t pulling back from live gatherings because AI gave them a shortcut. They’re leaning on them harder, at the exact moment general digital reach is getting harder to buy.

Both point to the same shift. Buyers aren’t distributing their trust more widely because AI gave them more places to look. They’re concentrating it, on fewer sources, chosen more deliberately, than they did before AI existed. Depth and presence are what’s earning that concentration. Everything else is losing ground to it.

The Verification Tax

Here’s the insight I didn’t anticipate, because it’s not a story about media at all. It’s a story about who’s making B2B purchase decisions with the least ability to catch AI when it’s wrong, and how much authority they’re being handed to act without checking.

Outsell asked buyers how they use AI relative to how well they already know the category they’re researching. 45% describe themselves as relatively new to the category, using AI to get oriented. 39% have general familiarity and use it to narrow options. Only 12% know the landscape well enough to use AI to move faster on specifics they already understand.

Trust runs backward from what you’d expect:

* Buyers new to the category: 93% rate their trust in AI high or very high, 55% at the maximum level. Asked if they’d trust it outright with no further checking, 56% said yes.

* Buyers who know the category well: 71% high or very high, just 25% at the maximum level, less than half the newcomer rate. Only 14% would trust AI outright with no checking.

Outsell has a name for the workaround buyers are building for themselves: the verification tax. 99% of respondents say they run some kind of validation process before acting on what AI gives them. 39% say they’re validating more than they did before AI entered the picture. The buyers who know the least about what they’re buying are the ones least likely to be paying that tax, at exactly the moment they should be paying it the most.

Each bar is a different population, not a slice of one total: category experts, category newcomers, and procurement buyers who are newcomers specifically. The percentage is the share of that group alone who’d trust AI outright, with no further checking.

The People Least Equipped to Check AI Are Gaining the Most Authority to Act on It

Break the newcomer numbers down by role and the risk gets sharper:

* New to the category they’re buying in: 55% of procurement respondents, versus 32% of end-user buyers.

* Trust AI outright, no further investigation, among newcomers: 67% of procurement respondents, versus 31% of end-user buyers.

* Say AI has led to meaningfully better decisions: 71% of procurement newcomers.

That gap isn’t a knock on procurement. Their job is to govern the process, control cost, and manage risk across categories they rotate through. But the people with the least domain depth are the ones extending AI the most unconditional trust, and they’re often the ones with signing authority.

Now put that next to where buying agents are headed:

* Already using an AI buying agent to identify, prioritize, and compare vendors, or to research the market outright: 49% of all respondents.

* Plan to adopt one within the next 12 months, among the 51% not yet using one: 85%.

* Delegating shortlisting decisions to an agent, now or within a year: 93% of B2B buyers, combined.

None of this means AI is giving buyers bad answers. Buyers rate AI’s usefulness at 4.3 or higher out of 5 at every stage of the purchase process, with no meaningful drop-off anywhere. The problem isn’t the tool. It’s that the people most likely to skip checking the tool’s work are gaining the most ground in how much of the decision they’re handing it.

What This Means

For CEOs and operators on the B2B buying side: Your own procurement process probably has this exact blind spot. Whoever on your team is newest to a category is also statistically the person most likely to trust AI’s output without checking it, and they may well be the person with budget authority to act on it. Build a verification step into any AI-assisted purchase above a material dollar threshold, especially where a less-tenured buyer is driving the research. Treat it as a governance question, not a technology question.

For B2B media, event, and marketing operators: Read the reliance numbers and stop there, and you’ll draw the wrong conclusion. The real finding isn’t that buyers need you more. It’s that they need you for something different than they used to. Top-of-funnel education, the job most editorial and content calendars are still built around, is losing ground to AI by design. The job that’s growing is validation: content and programming that helps a buyer confirm a shortlist AI already built, not content that helps them discover the category in the first place. If your content strategy, your event programming, and your sales pitch to sponsors are still oriented around top-of-funnel reach, you’re building for a job buyers have already handed to a machine.

Visibility in AI tools is table stakes now, the way SEO was a decade ago, but 89% of buyers treat it as one factor among several, not the only one. GEO gets a vendor onto the shortlist. It doesn’t win the deal. What wins the deal is the validation layer buyers are explicitly asking for: side-by-side comparisons, independent reviews, implementation reality that no vendor will volunteer. Build that, further downstream than your instincts are trained to go, and build it in the vertical where you already have the deepest relationships. Buyers are telling you directly they’re narrowing their trusted source list, not expanding it, and that the moment they need you has moved.

For investors: Two questions belong in every diligence checklist on a B2B media, information, or marketing asset from here forward. Has the company’s go-to-market spend actually shifted toward validation-stage content and away from top-of-funnel chase, or is it still fighting yesterday’s war? And does the business have genuine, defensible depth in a specific vertical, the kind Outsell calls the source of truth position, or is it competing on reach in a market that’s actively punishing reach? The businesses sitting on real depth are worth more than their current multiples suggest. The market hasn’t caught up to that yet.

AI didn’t take judgment out of the B2B buying process. It moved the moment judgment gets tested earlier, and made it easier to skip entirely. The buyers skipping it are telling us, in their own survey responses, that they know they’re taking a risk. The operators who build for that moment, not for the click that happens before it, are the ones who’ll still be standing when this shakes out.

My thanks to Outsell Founder and CEO Anthea Stratigos and Rob Keenan SVP and Lead Analyst for sharing the research.

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” is published by Tony Uphoff, Founder and Managing Partner of Uphoff Advisory: a strategic advisory practice for founders, CEOs, and investors in B2B information, marketing, and technology. The businesses that drive business.