Social Analytics

6.9% of ChatGPT Conversations Show Purchase Intent, SparkToro Panel Finds

6.9% Of ChatGPT Conversations Showed Purchase Intent In U.S. Panel
6.9% Of ChatGPT Conversations Showed Purchase Intent In U.S. PanelJESHOOTS-com / pixabay

SparkToro analyzed 41,961 U.S. ChatGPT conversations: 6.9% showed purchase intent and 87.9% of replies named no brand — bad news for ChatGPT ad hopes.

By Nathan Brooks

4 min read

At a glance

  • 6.9% of ChatGPT conversations in a 41,961-conversation U.S. panel showed purchase intent
  • 87.9% of ChatGPT responses named no brand, product or business
  • Panel: 12,321 people, Jan. 1–Oct. 29, 2026, data supplied by Verso
  • SparkToro's 2024 analysis of ~332M Google searches found 14.5% commercial intent
  • 131 heaviest users (1% of panelists) generated 21% of all prompts and were removed

Purchase intent showed up in just 6.9% of ChatGPT conversations in a U.S. opt-in panel analyzed by SparkToro — a figure that raises hard questions about where AI visibility budgets actually belong.

The data came from Verso, a tool that tracks users' ChatGPT history when they link their accounts. The sample covers 41,961 conversations from 12,321 people between Jan. 1 and Oct. 29, 2026, with most figures drawn from a sub-sample weighted to match U.S. demographics. In that weighted group, 87.9% of ChatGPT's responses did not name a single brand, product or business.

SparkToro co-founder and CEO Rand Fishkin concluded that the main reason AI visibility matters is Google's rollout of AI Overviews and AI Mode — not ChatGPT itself.

How often did conversations show buying intent?

The weighted panel splits into three purchase-intent buckets:

  • 4.9% of conversations: the person planned a purchase and was comparing options
  • 2.1%: questions about a specific brand, product or service
  • 93.1%: no purchase intention at all

"I suspect this is bad news for ChatGPT's advertising opportunities," Fishkin wrote.

He compared the result with SparkToro's December 2024 analysis of roughly 332 million Google searches, run on a U.S. browser panel from January 2023 through September 2024. Commercial intent — financial, business services or commerce — accounted for 14.5% of searches, and transactional queries for 0.69%. The categories don't match the ChatGPT purchase-intent chart, so the percentages can't be compared directly. Fishkin's broader definition covered 15% or more of searches.

How often did brands come up?

On the prompt side, 86.7% of all prompts didn't ask about any brand, product or company. Another 6.6% asked about several brands, and 6.7% asked about one. On the response side, 7.0% of replies mentioned multiple brands and 5.1% mentioned one — a combined 12.1% naming at least one.

The report's takeaways describe 12% of ChatGPT usage as part of the buyer journey, though its own purchase-intent chart puts buying-related conversations at 6.9%.

What did people actually use ChatGPT for?

Images and visual design led the category rankings at 17.6% of conversations. Information and entertainment came second at 14.4%, writing and communication third at 12.7%. Shopping, brands and local services made up 5.3%. Each conversation got one category based on its primary goal — output type outranked topic when someone asked ChatGPT to create something. Chart figures differ from the report's text by under two points per category; the numbers here follow the chart.

Compared with OpenAI's Signals data for September, the weighted panel leaned more toward image requests and how-to questions, while OpenAI's data ran heavier on writing and tech support. Fishkin called them fair to compare. OpenAI says its dataset covers only Free, Go, Plus and Pro accounts, excludes Codex and enterprise use, and reports global shares rather than U.S.-only ones.

How solid is the data?

Verso stores imported conversations and licenses them, de-identified, to research and business customers; co-founder Aaron Aknin supplied the data. The panel skewed young, female and lower-income, so 9,675 members with demographic data were weighted to the 2024 U.S. Census — no usage category moved more than 1.4 percentage points.

The 131 heaviest users, averaging over 33 prompts per active month, were cut before analysis. They represented about 1% of panelists but 21% of all prompts. The remaining weighted group averaged 9.9 prompts each between January and October.

Most panelists were recorded on mobile phones, which may undercount desktop and work-account use; API requests aren't included. Each conversation was coded with Claude using a written code book — titles, up to six prompts, opening response lines and product cards. A second coder recoded a random sample, agreeing on categories 90% of the time and brand tags 97% to 99% of the time. No agreement figure was reported for purchase intent specifically.

The panel reflects what people ask ChatGPT far better than how often they use it, the post notes.

What does this mean for budgets?

For PPC teams weighing ChatGPT Ads, the 6.9% share is a snapshot of a consumer panel with light ChatGPT use — not a count of purchase conversations across the platform. Fishkin questions whether tracking ChatGPT visibility is worth it unless you're also tracking Bing rankings and AI visibility. Until panel data covers desktop, work and API use, pair figures like these with your own referral and ad data before they touch a budget plan. Fishkin says he plans more research with Verso.

Source: searchenginejournal.com

Nathan Brooks

Show full bio

Staff writer covering companies, deals and strategy in Social platform news and policy at Reach Brief.

More from Nathan Brooks →

Related articles

« Previous articleNext article »