A new panel study of U.S. ChatGPT users found that only 6.9% of conversations contained any signal of someone looking to buy something. Even in those cases, most responses didn't name a specific brand. That's a useful reality check for any business owner being pitched on services to make their company "visible" inside AI chat tools.

What happened

Researchers analyzed a panel of real ChatGPT conversations from U.S. users to measure how often people use the tool with commercial intent โ€” meaning they're researching, comparing, or deciding on a purchase. The finding: fewer than 7 in 100 conversations showed that kind of intent. The rest were overwhelmingly about writing help, general information, coding, or casual questions.

Even more notable is what happened inside that small slice of purchase-related chats. Most of the time, ChatGPT's responses didn't surface a specific brand or product name at all. It tended to offer general advice, categories of options, or factors to consider rather than naming a specific company a user could then go buy from.

This matters because an entire industry has sprung up in the last year promising to help brands get mentioned by name when AI chatbots answer shopping-related questions. Agencies now sell packages under names like "AI search optimization" or "generative engine optimization," positioning them as the next must-have line item after traditional SEO. This panel data suggests the pool of conversations where brand-naming even happens is much smaller than the sales pitches imply.

The study doesn't claim purchase intent is zero or that it won't grow. It's a snapshot of current behavior, not a prediction. But it's one of the first attempts to quantify, rather than assume, how often ChatGPT functions as a shopping assistant versus a general-purpose tool.

Why it matters

This follows a familiar pattern in tech marketing cycles. When voice assistants like Alexa and Google Assistant became popular around 2017-2019, a wave of agencies sold "voice search optimization" services based on projections that voice shopping would explode. Actual voice-driven purchases stayed a tiny niche for years, and much of that spending produced little measurable return.

The current rush to optimize for AI chatbots is following a similar arc, built more on projected growth than present-day usage data. This study is one of the first attempts to put a real number on that gap, rather than leaving it to speculation.

What this means for small businesses

If a vendor is pitching you on getting your business mentioned by name inside ChatGPT responses, ask what share of your actual traffic currently comes from AI referrals. For most small businesses, that number is still small to nonexistent โ€” check your website analytics for referral traffic from chatgpt.com, perplexity.ai, or similar domains before committing budget to this channel.

That doesn't mean AI chat is irrelevant forever. Categories involving heavier research โ€” electronics, software, financial products, health services โ€” may see higher purchase-intent conversations than the overall average, since those are exactly the kinds of decisions people turn to a chatbot to think through. If you're in one of those categories, it may be worth monitoring more closely than a general retail or local-service business would.

For now, the safer move is to keep your existing SEO, Google Business Profile, and review presence solid โ€” those still drive the traffic that converts. Treat "AI visibility" spending as a small experiment, not a budget reallocation.

What to watch

Watch for follow-up studies tracking this percentage over time, since a rising trend would be a stronger signal than a single snapshot. Also watch whether OpenAI, Google, or Perplexity roll out dedicated shopping or checkout features inside their chat products โ€” that would change the incentive for these tools to name brands more often, and change the calculus for marketers.

The bottom line

Current data suggests ChatGPT is mostly used for things other than shopping research, and even when it is, it rarely names brands. Small businesses should size any investment in "AI search optimization" against their own referral data, not against vendor projections of where the market is headed.