Two weeks ago, we published the first-ever survey of AI enablement services buyers. One comparison from that survey keeps coming up in conversations since: in established professional services markets, price is one of the dominant competitive levers. In AI enablement, it is a distant third.

93% of buyers will disqualify a vendor for lack of AI credibility. 91% for lack of domain expertise. Price registers as a dealbreaker for 34%. Company size, for 9%.

On the positive side, domain expertise (86%) and proof of concept (56%) dominate positive selection criteria, and size/brand barely register (1%). This is extremely good news for AI founders: you're not competing on price or scale, you're competing on a credibility signal that buyers use as a proxy because they can't evaluate quality directly. The trick is to find ways to build trust quickly in a market that barely exists yet. 

The data makes more sense once you understand the structural problem buyers are navigating. AI enablement is new, and buyers' own experience in it is thin. This makes evaluating quality directly very difficult. Most buyers aren't in a position to run a rigorous capability assessment - they simply don’t know how to. So they buy the signal instead of the substance: they look for markers that let them infer quality rather than measure it.

Domain expertise, demonstrated results, and credibility indicators are all proxies for a judgment the buyer can't comfortably make themselves. Price is not a useful proxy for quality in an unproven market. A low price doesn't tell you whether a firm can actually do the work; it just tells you they're cheaper. That's not the risk buyers are trying to hedge.

The survey confirms this in how buyers describe their sourcing behavior. Vendor processes involve a small number of finalists. Primary sourcing runs through existing relationships and peer recommendations. Independent evaluation is uncommon. That means most of the competitive game is won (or lost) before a formal process begins. The shortlist forms in someone's professional network, not through a wide search and tender process.

Amplifying the difficulty is the fact that buyers are struggling to compare vendors on a like-for-like basis. It was the third most commonly cited reason for why buying AI enablement services is harder than buying any other professional services.

If you put the selection factors and the rejection reasons side by side, it’s basically the same story told twice.

On the way in, buyers say they’re choosing based on domain expertise (86%), proof of concept (56%), and reputation / credibility (51%). Price (11%) and company size (1%) barely come up.

On the way out, it flips but not really. Vendors get rejected for lack of AI credibility (93%) and lack of domain expertise (91%) far more than anything else. Price shows up at 34%, size at 9%.

So it’s not two different decisions. It’s the same question, just asked twice: do I trust you to do this? If the answer is yes, you’re in the mix. If it’s no, you’re gone before you even get a shot.

And then there’s the price. It’s surprisingly irrelevant. Only 11% of buyers pick a vendor because of it, and while 34% might walk away because of it, it’s still nowhere near the top reasons. It doesn’t really win you deals, and it doesn’t really save you either.

And yet a meaningful number of AI enablement firms are still leading with it. I suspect two things are driving that. One is confidence: selling on credibility means being able to defend it, and many early-stage founders aren't sure they can. The other is logic that isn't entirely wrong: price can get you a first engagement, and that first engagement can become the proof point you didn't have before. That path exists. But it requires a real upsell strategy and a deliberate plan to reposition on credibility once you're in. Without one, discounting becomes permanent positioning, not a temporary entry mechanism, and you've set your floor rather than opened a door.

Nine percent of buyers would disqualify a vendor for being too small or early-stage. Seventy-seven percent are currently working with a mix of firm sizes. Size, as a barrier to entry, is largely a non-issue – which makes it the wrong thing for new entrants to worry about.

Seniority creates meaningful texture here. Directors are more conservative: 29% work exclusively with large, established firms. VP and C-level buyers are considerably more open, with only 15% restricting to large-firm provenance. The new entrant window exists, but it's more accessible higher in the buying organization.

What's worked for founders trying to break in without an established name: borrowing credibility rather than waiting years to build it. For training firms, this typically means partnering with facilitators who bring recognizable institutional backgrounds: a former executive from a big name company, someone whose prior role does the trust signaling. For consulting and workflow implementation firms, vendor partnerships with established technology platforms create a similar effect. The logic in both cases is the same: you're not claiming credibility you don't have, you're demonstrating that credible people have chosen to affiliate with you.

That's a faster path to the proxy buyers are looking for than either undercutting on price or waiting for organic brand recognition to develop.

89% of buyers we surveyed are expanding spend with existing vendors. Only 7% anticipate running new vendor searches in the near term. If you can execute a few projects well, that will create a flywheel and pipeline for many years to come, cementing your position in the market. 

The standard playbook for scaling professional services is to spend your way into awareness: content marketing, paid distribution, conference presence, outbound, brand advertising. The assumption is that visibility creates familiarity, familiarity creates consideration, and consideration converts to revenue. We think that playbook doesn't work in AI enablement services. The path to being in the consideration set runs through people, not channels. And the path from consideration to selection runs through evidence, not messaging. Spend can amplify both of those things at the margins: you can use content to surface expertise, you can use distribution to get case studies in front of the right people, but it can't substitute for them.

The full survey is available here.

– Daria