ChatGPT usually leaves a SaaS product off its shortlist for one of five reasons, and you can check each one. Two sit on your own product and pricing pages. Two sit on review sites and third-party lists. The fifth is about the exact question your buyer types.
You've probably tried this already. You asked ChatGPT for the best tool in your category, and it answered with five names. None of them was yours.
We ran 25 software-buying prompts across ChatGPT, Perplexity, and Google AI Overviews, and logged every vendor named and every page cited. Below are the five checks ChatGPT's side of that data points to, in the order we'd run them. Each one comes with a test you can run this week.
Key Takeaways
- ChatGPT answers software questions with a shortlist of about five or six vendors. If you're not on it, buyers may never reach your site.
- For 58.9% of the vendors ChatGPT named in our audit, it cited no page on the vendor's own site.
- Qualifier prompts reshuffle the list: company-size, integration, price, and "alternatives" prompts named 38% to 54% new vendors across the three engines.
- ChatGPT cited vendors' product, pricing, and docs pages most, and Reddit never.
- Run five checks: qualifier fit, review profiles, readable product pages, public pricing, and list presence.
Why Isn't ChatGPT Recommending Your SaaS Product?
With search on, ChatGPT builds its shortlist from pages it finds at answer time, and most of them aren't yours. In our audit, it cited no page on the vendor's own site for 58.9% of the vendors it named. If you're missing, those pages probably don't describe you for that prompt.
Across our 25 prompts, ChatGPT retrieved 1,056 search results. It then cited about four sources per answer and named 5.6 vendors on average, 141 in total. The domains it retrieved most often were vendors' own sites and help centers, such as zendesk.com and docs.helpscout.com. That's one run on one day, from the US, logged out.
Four citation slots for 5.6 names explain part of the gap, since some named vendors can't get their own page cited. The rest depends on which pages win those slots. When a list wins instead of your product page, the list's wording probably decides whether you fit the prompt.
Absence isn't a verdict on your product. ChatGPT doesn't trial the software it names. A strong product described vaguely can lose the slot to a weaker one described clearly.
We split AI visibility into two layers. The cited source is the page ChatGPT links as evidence. The named vendor is the product it recommends. Most vendors ChatGPT named in our audit won the second layer without the first. Our pillar on GEO for B2B SaaS explains both layers across all three engines.
Some causes of absence apply to any brand, not only software; see why your brand doesn't appear in ChatGPT and AI answers. The five checks below are specific to software shortlists.
What Does Missing ChatGPT's Shortlist Cost You?
Missing the shortlist can cost you the evaluation before it starts. In our audit, ChatGPT named 5.6 vendors per software answer on average, and never more than 10. A buyer's follow-up questions will most likely be about those names. A vendor left off the first answer has to reach that buyer some other way.
G2 surveyed 1,076 B2B decision makers in March 2026. In its results, 69% of buyers chose a different software vendor than initially planned, based on AI chatbot guidance. The same survey found that 51% now begin software research with an AI chatbot more often than with Google. In April 2025, the figure was 29%.
That makes the shortlist a sales stage you can lose without knowing. A buyer who never sees your name never visits your site, so there's no visit to count and no referral to trace. We'd expect the pipeline to run thinner, with nothing in the data to explain why.
The reliable way to find out is to ask the questions yourself. Write the prompts your buyers would type, run them logged out, and note who ChatGPT names. That list is your real competitive set for AI search, and it may not match the one your sales team tracks.
Does ChatGPT Match You to Your Buyers' Qualifiers?
Many "why not us" cases may be qualifier gaps rather than brand gaps. Across the three engines in our audit, qualified prompts named many vendors the plain "best X" answer had left out. The new share was 38.5% for company size, 47.0% for integration, 38.2% for price, and 53.5% for alternatives.
A qualifier is the part of the prompt that narrows the category: company size, an integration, a price band, or a named competitor. "Best help desk software" and "best help desk software for a 10-person B2B support team" are separate contests, with partly different winners. A product missing from the first can still appear in the second.
The same numbers cut the other way. Because shortlists ran to about five or six names, new vendors meant some names from the plain answer dropped out. A place in the plain "best HR and payroll software" answer tells you little about the version "for a 30-person US startup." That longer prompt may be closer to what your buyers type.
To run this check, write 10 to 20 prompts with the qualifiers your buyers use. Include company size, your main integrations, your price band, and "alternatives to" your category leader. Two examples from our set are "best CRM for a 50-person B2B SaaS company" and "best help desk software that integrates with Salesforce." Run each one and record whether you're named.
Each qualifier likely pulls in different pages, such as integration directories or lists written for startups. The full mechanism is in How Do AI Engines Decide Which SaaS Vendors to Recommend?
Do You Have the Review Profiles ChatGPT Expects?
On ChatGPT, review sites were rarely a citation source. In our audit, ChatGPT cited a review site such as G2 or Capterra in only 2 of its 25 answers. Review sites made up 1.9% of its 103 citations. The check is simple: do complete, current profiles exist?
A low citation count doesn't mean reviews are irrelevant. Our reading is that they work one step removed from the answer. The lists ChatGPT cites may draw on review profiles, and a product without one gives them less to work with. That's an inference, not something our audit measured.
The engine matters here. In the same audit, Perplexity cited a review site in 14 of 25 answers, against ChatGPT's 2. If your review plan was built around AI visibility, check which engine it had in mind. We'd expect a review drive to show up in Perplexity's citations before ChatGPT's.
Start by searching G2 and Capterra for your product and for your three closest competitors. Confirm you have a profile on both, listed under the category your buyers search. Check that the description, pricing, and integrations match your current product page. Fix any gap before the next check.
Quoleady's June 2026 study supports the entry-requirement reading. It tested "dozens of" SaaS alternatives keywords but published no exact count. In its results, 100% of the tools mentioned in ChatGPT answers had reviews on Capterra, and 99% had reviews on G2. Yet review count and score correlated weakly or negatively with ranking position, from -0.21 to +0.02. Presence seems to matter, but a bigger review count is unlikely to lift your position. For the full picture, see Do G2 and Capterra Reviews Decide Whether AI Recommends Your SaaS?
Can ChatGPT Read What Your Product Does?
ChatGPT reads vendor pages more than any other source type, so your product page is often the evidence. In our audit, 65.0% of its 103 citations sat on vendor domains. And 51.5% of all its citations were named vendors' own product, pricing, or docs pages. That makes a vague product page a ChatGPT-specific weakness.
The check is whether a stranger could say what you sell from the top of your product page. It should name your category, the company size you serve, and your main integrations. A tagline like "the all-in-one platform for modern teams" gives ChatGPT nothing to match against a qualified prompt.
Then view the page source and search for your category name. If it's missing, your copy likely loads through scripts, and crawlers that read raw HTML may not see it. The full owned-page playbook is in How Do You Get Your SaaS Cited in ChatGPT and Perplexity?
Is Your Pricing Public?
A price-band prompt needs a price to match, and a "contact sales" page doesn't offer one. In our audit, price prompts named 38.2% new vendors across the three engines. On ChatGPT, the most-cited page type was vendors' product, pricing, or docs pages. We didn't test pricing pages directly, so treat this check as our reasoning.
One of our price prompts asked for the "best affordable CRM for an early-stage startup." If your pricing page says "contact us," we'd expect ChatGPT to favor a rival that publishes its price.
Open your pricing page in a private window and look for a starting price. If you can't publish one, state the pricing basis, such as per seat. The detail is in Does Your SaaS Pricing Page Affect Whether AI Recommends You?
Are You on the Lists ChatGPT Reads?
Check whether you're on the "best X" lists for your category, because ChatGPT cites them often. In our audit, 72% of its software answers cited at least one list page. Independent lists made up 29.1% of its citations, and vendors' own list pages another 8.7%.
On our five alternatives prompts, ChatGPT cited the market leader's own site in only one answer. If you're the leader, a rival's list may be describing you. If you're the challenger, that list is where you want to appear.
If your ChatGPT plan starts with Reddit, test that assumption. ChatGPT cited Reddit in 0 of our 25 answers. Of the 1,056 search results it retrieved, none came from Reddit.
Search "best [your category] for [your buyer]" and note which lists name you. The next step is in How Do You Get Your SaaS Into the "Best Tools" Lists AI Engines Cite?
Frequently Asked Questions
Why doesn't ChatGPT recommend my software?
The most likely reason is that the pages ChatGPT reads don't describe you for that prompt. In our audit, it cited no page on the vendor's own site for 58.9% of the vendors it named. Run five checks: qualifier fit, reviews, product pages, pricing, and lists.
Does ChatGPT use Reddit to recommend SaaS tools?
It didn't for the software-buying prompts in our audit. ChatGPT cited Reddit in 0 of 25 answers, and none of the 1,056 search results it retrieved came from Reddit. Its citations went mostly to vendors' product, pricing, and docs pages, and to lists.
Do G2 reviews help ChatGPT recommend my SaaS?
Having profiles looks like the baseline, not the edge. According to Quoleady's June 2026 study, every tool named in its ChatGPT answers had Capterra reviews and 99% had G2 reviews. Yet review counts and scores correlated weakly or negatively with ranking position. In our audit, ChatGPT cited a review site in only 2 of 25 answers.
How many vendors does ChatGPT recommend?
In our audit, ChatGPT named 5.6 vendors per software answer on average, 141 across 25 answers. It cited about four sources per answer. Qualified prompts reshuffled the list: across the three engines, company-size, integration, price, and alternatives prompts named 38% to 54% new vendors.
How do I check whether ChatGPT recommends my product?
Write 10 to 20 prompts with the qualifiers your buyers use, such as company size, integrations, price band, and "alternatives to" your leader. Run them logged out and record who ChatGPT names. Repeat the same prompts later before trusting any trend.
Conclusion
ChatGPT names about five or six SaaS vendors per answer, and for most of them it cites no page of their own. Run the five checks in order with 10 to 20 qualified prompts, logged out, and fix the first one you fail.
Agencies such as Kubnal Bridge specialize in running this baseline across engines and fixing the layer that fails. If you'd rather hand the work over, that's what we do.
About the Author
Kubnal Bridge Editorial Team. The Kubnal Bridge editorial team researches and writes about generative engine optimization and AI search visibility across ChatGPT, Perplexity, Google AI Overviews, Gemini, and Claude.
Sources
- G2 via PR Newswire, "New G2 Research: Half of B2B Software Buyers Now Start Their Research With AI Chatbots," 2026-04-15, https://www.prnewswire.com/news-releases/new-g2-research-half-of-b2b-software-buyers-now-start-their-research-with-ai-chatbots-302742807.html
- Quoleady, "Do G2 and Capterra Reviews Influence ChatGPT Rankings? [2026 Research]," 2026-06-04, https://www.quoleady.com/llmo-research/
- Kubnal Bridge, SaaS citation audit: 25 software-buying prompts across ChatGPT, Perplexity, and Google AI Overviews (first-hand data)