ChatGPT purchase intent is rare. In a United States panel of 41,961 ChatGPT conversations analysed by SparkToro using data from Verso, only 6.9% showed any intent to buy — 4.9% comparing options and 2.1% asking about a specific brand, product or service — while 93.1% showed none. Most answers do not name a brand either: in the weighted sample, 87.9% of responses mentioned no brand, product or business. Chasing ChatGPT mentions as a buying channel therefore reaches a small slice of a small slice. For link builders, the better focus is Google’s AI Overviews and AI Mode, where buyer-intent queries concentrate, and the citation-core sites those systems quote.
Those numbers come from a November 2026 study by Rand Fishkin of SparkToro, based on data supplied by Verso and reported by Matt G. Southern in Search Engine Journal; the original analysis is on SparkToro’s blog. This article explains what the study measured, why the 6.9% figure should cool the hype around ChatGPT as a buyer channel, and where link builders should focus instead.
What did the SparkToro and Verso study actually measure?
The study is based on an opt-in panel of United States ChatGPT users who agreed to share their chat history through Verso, which connects to a person’s ChatGPT account and analyses their history. The dataset covers 118,163 prompts across 41,961 conversations from 12,321 people between January 1 and October 29, 2026. Most headline figures come from a sub-sample weighted to match the demographic characteristics of the United States population.
Before the analysis, the researchers removed the 131 heaviest users, who averaged more than 33 prompts per active month. Those users were about 1% of panelists but sent 21% of all prompts. Each conversation got one label from a codebook of 34 intents grouped into 12 categories, using Anthropic’s Claude model. When someone asked ChatGPT to produce something — an image, an email, an essay — the output type counted over the topic, so a request to write a refund email was categorised as writing, not shopping.
Two details matter for anyone quoting the numbers. First, the panel reflects what people ask ChatGPT better than how often they use it: it is a consumer panel with light usage, most panelists were recorded on mobile phones, desktop and work-account use may be underreported, and API requests are not included. Second, the report’s text and chart figures differ by under two percentage points per category; the percentages here follow the charts, as Search Engine Journal does.
How much ChatGPT purchase intent is there, really?
In the weighted group, 4.9% of conversations involved a person planning to purchase while still comparing options, and 2.1% involved questions about buying a specific brand, product or service. Together that is 6.9%; the remaining 93.1% showed no intention to purchase. Fishkin’s summary was blunt: “I suspect this is bad news for ChatGPT’s advertising opportunities.”
For comparison, SparkToro’s December 2024 analysis of approximately 332 million Google searches found that 15% or more matched its purchase-intent definition. Commercial searches accounted for 14.5% and transactional searches for 0.69%. Search Engine Journal notes these categories do not match the ChatGPT analysis, so the percentages cannot be compared directly — but a much larger share of Google searching is tied to buying than of ChatGPT chatting.
Fishkin puts the scale gap plainly. He estimates Google gets about 5.5 times as many searches as ChatGPT gets prompts, and is roughly 2.8 times bigger in the share of queries with purchase intent — so people ask Google purchase-journey queries at more than 16 times the rate of ChatGPT. Those are estimates, not measured counts, but they frame the opportunity correctly: the buying conversation is overwhelmingly a Google conversation.

Why do so few ChatGPT answers name a brand?
The study tracked brands on both sides: whether people asked about brands, and whether ChatGPT’s answers named any. On the prompt side, 86.7% of prompts did not ask for brand, product or company information; 6.6% asked about several brands and 6.7% about a single brand. People mostly are not asking ChatGPT about brands.
On the response side, the pattern is similar. In the weighted group, 7.0% of responses mentioned multiple brands, products or companies and 5.1% mentioned one — a total of 12.1% mentioning at least one brand. The other 87.9% named no brand, product or business. ChatGPT mentions brands slightly more often than users ask for them, but the numbers stay small in both directions — and a brand can only be mentioned when the conversation is about something a brand could answer, which most ChatGPT conversations are not.
What do people actually use ChatGPT for?
The topic breakdown explains why purchase intent is so low. Images and visual design led at 17.6%, followed by information and entertainment at 14.4% and writing and communication at 12.7%. Shopping, brands and local services came eighth, at 5.3%:
- Images and visual design (17.6%): making or changing pictures and designing graphics.
- Information and entertainment (14.4%): fact-finding questions, movies, music, games, sports and news.
- Writing and communication (12.7%): help with texts, emails, social media, resumes and cover letters.
- Everyday life and admin: household fixes, recipes, trip planning and pet care.
- Health and wellness: symptoms, medications, diet, fitness and sleep.
- Money, work and business: banking, taxes, investing and job hunting.
- Shopping, brands and local services (5.3%): comparing products and finding nearby businesses.
Fishkin’s conclusion is that ChatGPT is a heavily consumer app, and that the competition with Google has been massively overblown by marketing press and influencers. People do not treat ChatGPT as a search engine. They treat it as a confidant, a creator tool, an entertainment product and an assistant — and most of what they ask has nothing to do with buying.
Is being mentioned in ChatGPT the same as being recommended?
No, and the distinction is where a lot of AI visibility reporting goes wrong. A mention means a brand’s name appeared in an answer. A recommendation means the AI presented that brand as a good choice for the person’s need. The study measures mentions, not recommendations, and the two are very different things for a buyer.
Consider what a mention can look like in practice. A brand can be named in a list of five options, as an example of what not to buy, or in passing while the answer recommends something else. None of those is a recommendation, and none moves a buyer. A report that counts mentions without reading the answers can look healthy while the brand is never actually recommended to anyone.
There is also a numbers trap here. Fishkin notes that 12% of ChatGPT usage is part of the buyer journey — but that reflects the 12.1% of responses mentioning a brand, not the 6.9% of conversations showing purchase intent. Mentioning a brand and intending to buy are different measurements, and the gap between them is where inflated claims live. Before acting on any AI visibility number, read the underlying answers: which prompts produced the mention, what the answer said, and whether the person was near a purchase decision.
ChatGPT vs Google AI surfaces: where does buyer visibility actually concentrate?
Fishkin’s central conclusion is that AI visibility matters a huge amount — but not because of ChatGPT. In his words, it is Google, and Google’s AI Overviews and AI Mode, that make AI visibility so important. The comparison below summarises why.
| Factor | ChatGPT | Google AI Overviews and AI Mode |
|---|---|---|
| Share of interactions with purchase intent | 6.9% of conversations | 15%+ of searches matched the purchase-intent definition (SparkToro’s 2024 Google analysis) |
| How often answers name a brand | 12.1% of responses mention at least one brand | AI Overviews cite sources for queries users already ask Google |
| Overall volume | Smaller; Google gets about 5.5 times as many searches as ChatGPT gets prompts (Fishkin’s estimate) | Larger; purchase-journey queries estimated at more than 16 times the rate of ChatGPT |
| Where buyers compare options | Rarely; shopping is eighth at 5.3% of conversations | Routinely; commercial and transactional queries are a core part of search |
| What link building can influence | Indirectly, through sources ChatGPT may draw on | Directly, through the cited sources and pages AI Overviews quote |
The practical reading is simple. ChatGPT visibility is worth tracking only if you are also tracking keyword rankings and AI visibility on Bing — Fishkin’s own test. If Bing is too small to be worthwhile but ChatGPT is treated as worthy of time and budget, he calls that intellectually dishonest. Google’s AI surfaces sit on top of the search behaviour where buyers already compare options, so they deserve the larger share of attention. Our analysis of Google search volume versus ChatGPT reaches a similar conclusion, and our guide to where Google rankings and AI recommendations overlap shows how the two connect.
Where should link builders focus instead?
If ChatGPT mentions are a small prize, the work that moves buyer visibility looks different. Link building for AI search is not about getting named in a chatbot; it is about becoming a source AI systems quote when buyers ask comparison and purchase questions. Five priorities follow, in order.
- Win visibility in AI Overviews first. Optimise the pages that already rank for buyer-intent queries so they are also cited in AI Overviews and AI Mode. Answer-first openings and question-based headings make a page easy for an AI system to quote accurately. Our breakdown of share of recommendations in AI search explains why recommendation share, not raw mention counts, is the number worth watching.
- Build links on citation-core sites. AI systems quote a relatively small set of authoritative sources again and again. A link on a site AI Overviews regularly cite is worth far more than a mention on a site no AI system ever quotes. Prioritise real, topically relevant publications, and check whether those sites appear as cited sources in AI answers for your target queries.
- Cover buyer prompts, not just keywords. List the questions buyers ask when comparing options: “best X for Y”, “X vs Y”, “how much does X cost”, “is X worth it”. Then create or improve pages answering each one directly, with the answer in the first 100 to 150 words. These are the prompts where a citation can turn into a purchase — and the ones most AI visibility tools never test.
- Earn brand mentions where buyers compare options. Guest posts, expert quotes and original data on the sites buyers read while comparing options do double work: they build the links search engines use, and they put the brand in the sources AI systems draw on. A guest post on a niche-relevant site AI Overviews cite beats content aimed at ChatGPT alone.
- Measure recommendations, not mentions. Track which prompts cite your pages, what the answer says about your brand, and whether the prompt came from a buyer. A rise in mentions with no rise in recommendations is noise; a rise in recommendations on buyer prompts is signal. Keep that separate from referral traffic, which is a different number.
None of this means ignoring ChatGPT entirely. The 6.9% is small, but it is not zero, and the 12.1% of responses that mention a brand show real buyer-journey activity. The mistake is treating it as the main event. Put the larger share of effort where buyers already are — Google’s AI surfaces and the sources they cite — and treat ChatGPT as a secondary channel, not the headline goal.

When should you hire a service like VonixCo?
Doing this work in-house makes sense when you have a clear list of buyer prompts, relationships with niche publications, and someone who can place guest posts and track AI citations week after week. It stops making sense when placements stall, the sites you reach are not the ones AI systems cite, or nobody has time to keep it going.
That is the point at which hiring a service like VonixCo is worth considering. VonixCo is a niche-relevant guest posting and SaaS link building service based in Lahore, Pakistan. Its services are SaaS guest posts, niche guest posts, link building and guest post content writing, with plans at $500, $950 and $1,800 per month, content included and a live link report with every placement. A service like this fits best when you want placements on real, niche-relevant sites — including citation-core sites AI Overviews quote — without building outreach yourself. Contact: info@vonixco.com and WhatsApp +92 340 0096621.
Be careful about what any service can honestly promise. No link building service can guarantee that Google or an AI system will rank a page or recommend a brand. What a good service can do is place your content on relevant, authoritative sites, keep the work transparent with a live link report, and focus on the pages and prompts where buyers compare options.
What are the limits of this data?
This study is the best public look at ChatGPT usage so far, but its limits should stay attached to the headline number. The panel was opt-in and skewed toward younger people, women and people at lower income levels; figures were weighted to 2024 United States Census targets, and no usage category varied by more than 1.4 percentage points when weighted. It is also consumer-heavy: it excludes API requests, and desktop and work-account use may be underreported because most panelists were recorded on mobile phones.
Conversations were coded by an AI model from a written codebook, with a second coder independently recoding a random sample; the two agreed on categories 90% of the time and on brand tags 97% to 99% of the time, but the report gives no separate agreement figure for purchase intent. Text and chart figures differ by under two percentage points per category. Treat 6.9% as a measured share of one consumer panel, not a count of purchase-related conversations across all of ChatGPT — and pair it with your own referral and search data before putting it into a budget plan.
Frequently asked questions
What is ChatGPT purchase intent?
It means a conversation shows the person is considering buying something. In the SparkToro and Verso study, that covered planning a purchase while comparing options (4.9% of conversations) and asking purchase-related questions about a specific brand, product or service (2.1%). Together, 6.9% showed purchase intent.
How often do ChatGPT answers mention a brand?
In the weighted sample, 12.1% of responses mentioned at least one brand — 7.0% mentioned several and 5.1% mentioned one. The other 87.9% named no brand at all. On the prompt side, 86.7% of prompts did not ask for brand information.
Why is ChatGPT purchase intent so low?
Most people use ChatGPT for things unrelated to buying. Images and visual design led at 17.6%, followed by information and entertainment at 14.4% and writing and communication at 12.7%. Shopping, brands and local services ranked eighth at 5.3%.
Does AI visibility still matter if ChatGPT purchase intent is only 6.9%?
Yes, but mainly because of Google rather than ChatGPT. Fishkin concludes that Google’s AI Overviews and AI Mode are what make AI visibility important, since buyer-intent queries concentrate on Google. Track ChatGPT visibility only alongside keyword rankings and AI visibility on Bing.
What is the difference between a ChatGPT mention and a recommendation?
A mention means a brand’s name appeared in an answer; a recommendation means the AI presented that brand as a good choice for the person’s need. A brand can be mentioned in a list, as an example, or in passing without being recommended. Read visibility reports answer by answer, not as a single mention count.
Where should link builders focus for AI search visibility?
Focus on Google’s AI Overviews and AI Mode, the citation-core sites those systems quote, and pages that answer buyer comparison questions directly. Build links and brand mentions on real, niche-relevant publications AI systems cite, cover the questions buyers ask when comparing options, and measure recommendations on buyer prompts rather than mention counts.
The 6.9% figure is not a reason to abandon AI visibility work. It is a reason to aim it correctly. Most ChatGPT conversations are not about buying, most answers name no brand, and the buying conversation still happens overwhelmingly on Google. Focus on AI Overviews, citation-core sites and genuine buyer questions — that is where the buyers are.