Zero-click buying in ChatGPT: what it means for B2B websites and lead capture

Zero-click started as a search term for results pages that answer the question so well the user never clicks. In B2B buying, it now describes something larger: buyers who learn about the category, compare vendors and form a shortlist inside AI answers, then contact one or two vendors at the end.

Forrester’s January 2026 research found that 94% of business buyers use AI in their buying process and that generative AI is named the most important information source twice as often as any other, including vendor websites. Forrester expects buyers to spend more of their process with answer engines and less time engaging with vendors directly.

The best independent evidence on clicks comes from Google, not ChatGPT. In a study of 900 US adults’ browsing in March 2025, Pew Research Center found users clicked a traditional search result in 8% of visits when an AI summary appeared, against 15% when none appeared. Only 1% of visits included a click on a link inside the summary. Google disputed the methodology.

ChatGPT now reports website clicks in Business Hub, which will give companies their first official view of how often a ChatGPT answer leads to a click. OpenAI notes those clicks do not confirm visits or conversions. Our launch guide to ChatGPT Business Hub explains how to claim the profile and see the number.

Our view is that fewer clicks are not the main risk. The bigger risk is being absent from the answer, because a buyer who never sees your name has no reason to search for you later.

Not for enterprise B2B in the near term. OpenAI’s business agents, available at launch to eligible ecommerce and online service businesses in the US, answer questions from the information you provide, cannot complete purchases, bookings or order lookups, and link people to your website. Enterprise purchases also involve buying groups, security reviews, procurement and contracts that no assistant handles today.

Our view is that B2B zero-click is a research shift, not a checkout shift. The visit you lose is the early, curious one. The visit you keep is the late, decided one, and it is worth more.

Watch for change, because the direction is set. If OpenAI extends business agents beyond the US or adds actions such as booking a demo, B2B sites will need the same approved facts in a feed as on the page. Keeping product, pricing and policy facts in one structured source now makes that step easier.

Publishing pricing and comparisons is uncomfortable for many enterprise teams. Our view is that if you do not publish the answer, an assistant will build one from review sites, competitors and old forum threads. Our article on publishing pricing for AI search covers how to do it without losing negotiating room.

1. Add an AI assistant option to the how-did-you-hear field on every demo and contact form, and ask the same question in the first sales call.
2. Create an AI channel in GA4 for chatgpt.com, perplexity.ai, gemini.google.com, copilot.microsoft.com and similar referrers, and carry it to the CRM.
3. Watch branded search and direct traffic to deep pages; buyers who read an AI answer often search your name later.
4. Shorten the path from pricing, comparison and case study pages to a sales conversation, and test it with conversion rate optimization.
5. Report AI-referred and AI-influenced pipeline monthly next to Business Hub website clicks.

We recommend making these changes, with our conversion rate optimization team, in the first month of a GEO program, because better AI visibility is wasted if decided buyers cannot convert.

Most B2B websites were built to attract early-stage visitors and nurture them. In a zero-click journey, the early-stage reading happens in the answer, so the site has to serve two readers: the assistant that summarises it and the decided buyer who arrives later.

• Every key page answers its main question in the first two sentences, in plain words an assistant can quote.
• Facts are consistent everywhere: company description, categories, locations, integrations and customer types match across the site, Business Hub, directories and review profiles.
• Pricing, implementation time, contract terms and security answers are published, or at least the model and the ranges.
• Proof is short and specific: customer, industry, problem and result, with permission to quote.
• Comparison and alternatives pages exist for the competitors you meet most often, written factually.
• Content that only existed to capture early-stage traffic is reviewed, merged or retired, so assistants find one strong page instead of five weak ones.
• Every page carries a clear next step for a decided buyer: a demo, a pricing call or a technical review.

This is also where many enterprise sites lose AI visibility today: key facts live in PDFs, behind forms or in scripts that crawlers do not render. A technical check of what AI crawlers can read is part of the first month.

Stop judging content by traffic alone. Judge it by whether assistants cite it and whether the buyers who arrive convert. Fund the pages that answer the late-stage questions, publish proof that is easy to quote, and measure AI influence in your CRM. Our generative engine optimization service covers the visibility side.

If you want to see how a zero-click journey looks for your category, talk to our team. We will walk through it in a free 30-minute pipeline plan.