AI Search Is Changing B2B Discovery: What to Do About It

The most important reader of your website no longer has a pulse. Before a buyer ever fills out a form, an AI has read your site, decided what you do, and either put you on the shortlist or quietly left you off it. I found this out the uncomfortable way, on a client I was proud of.

I pasted a client's homepage into ChatGPT, and the answer stung

A few months ago I ran a small test on a client we had worked with for years. Good firm. Real specialty. Clients who rave about them. I pasted their homepage URL into a chatbot and asked a plain question: based on this page, what does this company do, and who does it help?

The answer came back as one calm, polished, beige paragraph. Nothing was wrong. Nothing was memorable. It could have described forty other firms. This was a company that had won awards in its category and had a specialty its best clients could name in one sentence. None of that survived the read.

Here is the part that matters. Your buyer's first reader is no longer a human. It is a model that scans the words on your page and checks whether they answer the buyer's specific question. If your homepage gives a vague answer to a specific question, the AI passes that vague answer along, and the buyer never shows up. You do not get a "no thanks" email. You just get fewer of the right calls, and you wonder where the pipeline went. That is silent rejection, and it is happening to firms that have no idea it is happening.

We rebuilt that client's positioning, words first and design second. Then I ran the same test. This time the summary named the industry they serve, the outcome they deliver, and the buyer they are built for. Within a few weeks, buyers started quoting the new positioning back on sales calls: we saw that you specialize in this, and that is exactly our problem.

Your buyers really are starting with the machine

This is not a fringe behavior. As of early 2026, 51 percent of B2B software buyers say they now start their research with an AI chatbot more often than with Google, up from 29 percent less than a year earlier, and 71 percent now rely on AI chatbots somewhere in that research (G2, 2026).

And it is not just where they start. It is what they decide. In G2's global buyer study, generative AI chatbots became the single most influential source for building a shortlist, at 17.1 percent, ahead of software review sites at 15.1 percent and vendor websites at 12.8 percent (G2, 2025). Sit with that. An AI's recommendation now outweighs your own website in shaping who makes the list.

The most uncomfortable stat for anyone who assumes brand awareness protects them: 33 percent of buyers said they purchased from a vendor they had never heard of before, on the strength of what a chatbot told them (G2, 2026). A third bought from a stranger the AI vouched for. That is either your biggest threat or your biggest opening, depending on whether the answer engine can find a clear, credible reason to name you.

For an operator, this has a price you can estimate. If you run a five million dollar B2B service firm, and your site should be driving fifteen to twenty-five percent of inbound, a site the AI cannot describe is not neutral. It is quietly forfeiting hundreds of thousands of dollars in pipeline a year to firms the machine found easier to explain.

One honest caveat before we go further: buyers still want a human. Gartner found that 69 percent of B2B buyers turn to a sales rep to validate what an AI told them (Gartner, 2026). The AI builds the list. People still close. Getting named is necessary now. It was never sufficient.

Meanwhile, the click you used to earn is disappearing

Even when you do rank, the visit is worth less than it was. When Google shows an AI summary, people click a traditional link on 8 percent of those searches, compared with 15 percent when there is no summary, and they click a link inside the summary itself just 1 percent of the time (Pew Research Center, 2025).

The pages that used to win feel it most. Ahrefs measured the top organic result losing 34.5 percent of its clicks when an AI Overview appears, as of April 2025, and their updated read put that closer to 58 percent by December 2025 (Ahrefs, 2026). The broader backdrop: as of early 2026, roughly 68 percent of U.S. Google searches end without any click to the open web at all (SparkToro, 2026).

Ranking first matters less when a machine-written answer sits above you and settles the question on the spot. The value is moving from being ranked to being the source the answer is built from.

This is a positioning problem, not a plugin problem

The reflex is to treat this as a technical chore. Bolt on some FAQ schema, add a few question-shaped headings, call it "answer engine optimization," and move on. That misreads the shift. Schema cannot fix vague copy. It can help an AI read a page whose words are already clear, and it does nothing for a page whose words are wrong.

Here is the reframe I keep coming back to.

Ranking got you found. Citability gets you named.

What changed is where the decision forms. Buyers assemble a shortlist, and often a preference, before they ever talk to you, and increasingly before they ever land on your site. So the real question is no longer how do we rank. It is this: when an AI or a person is deciding who belongs on the list, is there a clear, consistent, believable case to point to. That is a positioning and content problem first, and a technical one second.

There is real research under this. In the peer-reviewed paper that named the field, adding verbatim quotations raised a page's visibility in AI answers by roughly 43 percent, adding statistics by roughly 33 percent, and citing credible sources by roughly 28 percent, while old-school keyword stuffing actively hurt (Aggarwal et al., "GEO: Generative Engine Optimization," KDD 2024). And Pew found that 88 percent of AI summaries cite three or more sources, so the goal is not to be the one answer. It is to be one of the several sources the machine trusts enough to name.

What to actually do about it

None of these moves are new. AI search just raised the price of ignoring them.

Can an AI recommend you?

Score your own firm. Tick each one you can honestly answer yes, then count your checks against the guide below. Do the first one for real: open a chatbot, paste your homepage URL, and ask what your company does and who it is for.

• Paste your homepage into an AI and ask what you do and who you serve. The answer comes back specific and correct.
• One plain sentence names your category and your buyer, in their words, not your org chart's.
• Your proof lives as text: specific numbers, named outcomes, quotes an AI can lift, not locked inside images.
• Your positioning reads the same on your site, on LinkedIn, and in the directories buyers check.
• Someone other than you (a review, a podcast, a guest article) has described what you do, in their own words.
• Your key pages were updated in the last twelve months, and they say so.

5–6

an AI can already sell for you.

3–4

a coin flip. Fixable, but leaking.

0–2

you are being quietly skipped.

Run the full Scorecard→

Don't take my word for it

The day I wrote this, I ran the same 20-second test on our own homepage. Here is the answer a chatbot gave, in plain terms: a boutique agency that builds conversion-focused B2B websites in eight weeks, backed by a guarantee to double conversion within ninety days, for B2B firms roughly between two and twenty million in revenue whose sites underperform. It pulled the real client results, named the buyer, and even quoted our headline back.

I am not showing you that to brag. I am showing you the bar. That is what a page reads like when the words are right, and it is the difference between an AI selling for you and an AI skipping you. It is also reproducible: you can run the exact test on us, and on yourself, in the next five minutes.

The approach behind it is not a trick. Across the rebuilds we have run, the pattern is consistent. A firm scoring in the thirties on our own diagnostic climbs into the high sixties or seventies, and the visitor-to-qualified-lead rate roughly doubles, or better, on the same traffic (anonymized client engagements, results vary). Clear positioning is what a human buyer rewards, and it turns out to be exactly what the machine rewards too.

The honest part

Two things are true at once, so let me hold both.

I am not an AI skeptic. I use these tools every day, and they have genuinely changed how I work and think, not just how much I get done. So when I say your buyers are leaning on them, I am describing a pull I feel myself.

And I am not selling panic. The biggest numbers in this space are contested, and you should know it. Gartner's forecast that search volume will fall 25 percent by 2026 is a scenario model, not a measurement (Gartner, 2024). Google disputes the Pew study's methodology. The foundational GEO research ran much of its testing on a simulated engine, not today's live ChatGPT or Gemini. And in absolute terms, AI referral traffic is still small: SparkToro clocked Google's own AI Mode at well under one percent of searches in early 2026.

But you do not need the forecast to be exactly right. You need it to be directionally true, and every independent source agrees on the direction. The move is not to abandon search for a shiny new acronym. It is both, and: keep earning rankings through solid SEO and content, and build the brand and clear positioning that make you the obvious answer when the decision happens somewhere you cannot see. Treat your brand as a retrieval signal for machines, not only a trust signal for humans. And when the AI does send you a higher-intent visitor, make that click count with a page built to convert it, which is the whole job of a conversion system.

That is the architect's version of SEO. It survives whichever way the numbers land.