AI Search Is a Lead Source With No Auction for the Answer

Buyers increasingly ask AI assistants such as ChatGPT, Claude, Gemini and Perplexity which suppliers to consider. The companies named in those answers win attention before a buyer ever visits a search results page. You cannot bid to be named in the answer itself. You can only be the company the evidence describes most clearly.

Why is AI search a lead source?

Because buyers now ask AI assistants the questions they used to type into search engines, including "which suppliers should I consider for this?". The assistant answers with a short list. A company named there gets a warm introduction to a buyer who is actively looking, without a click on an ad or a page of results.

The audience is large. According to TechCrunch's report of OpenAI's figures, ChatGPT reached 900 million weekly active users in February 2026, up from 800 million in October 2025. Business buyers are a fraction of that, but they are a growing fraction, and they are asking commercial questions.

At the same time, traditional search sends fewer clicks. According to Pew Research Center's study of 68,879 searches by 900 US adults, users clicked a traditional result in 8% of visits when Google showed an AI Overview, against 15% when it did not. Only 1% clicked a link inside the summary. More of the decision happens inside the answer.

Gartner forecast in 2024 that traditional search volume would fall 25% by 2026 as AI assistants took over queries. Treat that as a forecast rather than a measurement. Still, the direction matches what Pew observed.

Is there really no auction?

Not for the answer itself. ChatGPT began testing labelled ads in February 2026 for some US users on free tiers, shown separately below the answer and marked as sponsored. OpenAI says ads do not influence the answers ChatGPT gives. So a company can buy an ad beside the answer, but not a place in it.

That distinction is the point. In paid search, the top of the page goes to whoever bids most, and a competitor can push your cost per click up overnight. Inside an AI answer, the recommendation is meant to reflect what is useful to the person asking. OpenAI's January 2026 post on its approach to advertising calls this "answer independence".

Two caveats keep this honest. First, this is a vendor's stated policy, and policies change, so check it periodically. Second, paid tiers of ChatGPT carry no ads at all, and other assistants have their own approaches. Either way, the organic recommendation is not something you can bid on today.

How do buyers use AI to choose suppliers?

Buyers use AI assistants to understand a problem, compare approaches, build a shortlist and check what others say about each option. A typical sequence starts broad ("how do companies like mine solve this?") and narrows ("which providers suit a 50-person business in Australia?"), with follow-up questions about price, fit and reviews.

That sequence matters for how you show up. The early, problem-level questions are answered from explainers and guides. Later, shortlist questions draw on comparisons, reviews and directories. Finally, the checking questions draw on what customers and independent sources say. A company needs to be present at each stage, in sources the assistant trusts.

Why does this favour smaller companies, and when does it not?

It can favour smaller, specialist companies, because a precise answer to a specific question beats a big brand's generic page. But research also finds AI search leans heavily on established, third-party sources, which can favour big brands with more coverage. Specialists win when credible sources describe them clearly.

A 2025 University of Toronto study of AI search engines found "a systematic and overwhelming bias towards earned media", meaning third-party, authoritative sources, over content brands publish about themselves. The authors also warned of a "big brand bias" that can disadvantage niche players.

So the opportunity is real but conditional. A specialist that is clearly described across reviews, directories, partner pages and independent articles can be named alongside, or ahead of, larger competitors for a specific question. A specialist that exists only on its own website usually is not named at all.

What earns a recommendation from an AI assistant?

Clear, consistent, credible evidence. That means one plain description of what you do and who for, used everywhere; third-party coverage such as reviews, directory listings and articles; content that answers buyers' real questions with specific, sourced facts; and pages that AI crawlers can read. There is no submission form and no trick.

The essentials:

1. One description. The same name and one-line description of what you do and who it suits, across your site and every profile.
2. Third-party evidence. Reviews, directory listings, partner pages and articles that describe you accurately.
3. Answers to real questions. Explainers and comparisons that address what buyers ask, with honest trade-offs.
4. Readable pages. Content that does not depend on scripts to appear, and a robots file that admits AI crawlers.

The guide to getting recommended by ChatGPT works through each in detail.

How do you know if AI search is working?

Ask the assistants the questions your buyers ask, record which companies they name and which sources they cite, and repeat monthly. Add a "how did you hear about us?" question to your forms, and watch for AI assistants in your referral data. Change takes months, so track the trend rather than a single result.

Answers vary with phrasing and over time, so test several versions of each question across ChatGPT, Claude, Gemini and Perplexity. The guide to measuring AI visibility sets out a simple scoring method.

Skein measures how often the major assistants recommend a client, monthly, alongside pipeline and search demand; the outcomes page shows what is tracked. That suits a B2B business whose buyers research before they talk to sales. A business that wins mainly through referrals and repeat customers may see less from it.

Frequently asked questions

Can you get leads from ChatGPT?

Yes. Buyers ask ChatGPT and other assistants which suppliers to consider, and companies named in the answer get high-intent attention. Being named depends on clear, consistent descriptions of your business and credible third-party evidence, not on payment. Ask new leads how they found you to track it.

Can you pay to appear in ChatGPT answers?

Not in the answer itself. ChatGPT began testing labelled ads in February 2026, shown separately below answers for some US users on free tiers. OpenAI says ads do not influence the answers. Paid plans carry no ads. Being recommended within the answer depends on evidence, not spend.

Is AI search replacing Google?

Not replacing, but taking a growing share of research. Pew Research Center found Google users clicked traditional results less often when an AI Overview appeared, 8% of visits against 15%. Gartner forecast a 25% fall in traditional search volume by 2026, though that was a prediction, not a measurement.

How do AI assistants choose which companies to recommend?

They draw on what they can read about a company, especially credible third-party sources such as reviews, directories and articles. A 2025 University of Toronto study found AI search shows a strong bias towards earned media over brand-owned content. Clear, consistent descriptions across many sources help most.

How do you track leads from AI search?

Test the questions your buyers ask across several assistants each month and record which companies are named. Add a "how did you hear about us?" field to forms, and check referral data for AI assistants. Track the trend over months rather than reacting to a single result.

Related reading