To get your SaaS cited in ChatGPT and Perplexity, build different pages for each engine. ChatGPT quotes vendors' own pricing, integration, and help pages. Perplexity quotes lists, and many of those lists sit on vendors' own blogs.
This guide focuses on ChatGPT and Perplexity and sets out the work in order. Let the crawlers in, build the pages ChatGPT quotes, and publish the lists Perplexity reads. Then earn the off-site mentions that get you named, and check that it's working.
Key Takeaways
- ChatGPT and Perplexity cite different pages. For the same prompt, the domains they cited overlapped by 5.5% on average in our audit.
- ChatGPT's most-cited page type was the vendor pricing page, at 30.1% of its citations.
- Perplexity cited no pricing pages, but a vendor's own "best X" or alternatives list appeared in all 25 of its answers.
- Start by letting the search crawlers in: OpenAI says sites that block OAI-SearchBot won't be shown in ChatGPT search answers.
- Most vendors are named without their own site cited, so off-site lists and reviews still shape the shortlist.
What Does It Take to Get Your SaaS Cited by Both Engines?
Treat ChatGPT and Perplexity as two engines, because they read different sources. In our audit, ChatGPT cited 4.1 sources per answer, 65.0% of them on vendor domains. Perplexity cited 19.9 per answer, and 81.9% were lists. For the same prompt, their cited domains overlapped by only 5.5%, while the vendors they named overlapped by 44.4%.
That gap matters because of two separate outcomes. You're cited when an answer links to your page as a source. You're named when the answer recommends your product. An engine can name you while citing someone else's page. Our pillar on GEO for B2B SaaS explains why those two layers need separate work.
The general mechanics, such as answer-first writing and freshness, apply to every industry. We cover them in our guide on how to get cited by ChatGPT, Perplexity, and Google AI Overviews, so this guide sticks to SaaS pages.
Our figures come from a single run. We used 25 prompts across five software categories. Each category was asked five ways: plain "best X", then with a qualifier: company size, integration, price band, or "alternatives to" the market leader. ChatGPT ran logged out from the US with search on. Perplexity ran through its Sonar API, which is close to the consumer app but not identical. Treat category-level differences as indicative.
Work through five steps, in this order:
- Let the search crawlers reach your pages.
- Find which of your pages ChatGPT cites.
- Build pricing, integration, and help pages it can quote.
- Publish fair "best X" lists that Perplexity reads.
- Earn the off-site mentions that get you named.
Step 1: Can ChatGPT and Perplexity Reach Your Pages?
Check crawler access before you write anything. Search crawlers and training crawlers are different bots, and a SaaS site can block both with one rule without meaning to. If the search bots can't fetch your pricing and docs pages, the rest of this guide has little to work with. Start with the table below.
| Crawler | Owner | What it does | Ran JavaScript in Vercel's December 2024 test? |
|---|---|---|---|
| OAI-SearchBot | OpenAI | Surfaces sites in ChatGPT's search features; sites that block it aren't shown in ChatGPT search answers | No |
| GPTBot | OpenAI | Crawls content that may be used to train OpenAI's models; separate from search | No |
| ChatGPT-User | OpenAI | Visits a page when a user's question in ChatGPT triggers it; "robots.txt rules may not apply" | No |
| PerplexityBot | Perplexity | Surfaces and links sites in Perplexity search results; Perplexity recommends allowing it to ensure your site appears; not used to train foundation models | No |
| Perplexity-User | Perplexity | Fetches a page when a user's question triggers it; "generally ignores robots.txt rules" | Not tested |
Sources: OpenAI and Perplexity crawler documentation; Vercel, "The rise of the AI crawler," December 2024.
The table matters because these bots do different jobs, yet a single blanket rule in robots.txt or a firewall treats them all alike. A security team that blocks "AI bots" to stop model training may also block the bots that put you in answers. That's the mistake to check for first. OpenAI's documentation is explicit about the cost. Sites opted out of OAI-SearchBot will not be shown in ChatGPT search answers, though they can still appear as navigational links. GPTBot, by contrast, crawls content that may be used in training, so blocking it and allowing OAI-SearchBot are separate decisions.
Allow OAI-SearchBot and PerplexityBot even if you decide to block GPTBot. Whether to let your content train models is a policy choice. Whether to appear in answers shouldn't be decided by the same line of robots.txt.
Next, look past robots.txt. CDN and bot-protection rules can block a crawler that robots.txt allows, and the block won't show up in the file. Ask whoever runs your CDN or firewall which bot rules are active. Then fetch your pricing page with each search crawler's user agent and confirm the server returns the full page, not a challenge screen.
Last, check your rendering against the JavaScript column in the table. View the raw HTML of your pricing and docs pages, not the rendered page. Your plan names, prices, limits, and setup steps should be in that HTML as text, not loaded by scripts afterward.
Step 2: Which of Your Pages Does ChatGPT Cite?
Your pricing page, more than any other. It was ChatGPT's most-cited page type in our audit, with 31 of its 103 citations, or 30.1%, across 12 of 25 answers. Product or feature pages were 10.7% of citations, integration pages 6.8%, and help-center or docs articles 5.8%.
The pattern follows the prompt. Pricing pages were cited on 4 of 5 company-size prompts, 4 of 5 alternatives prompts, and 3 of 5 price prompts. On plain "best X" prompts, they appeared in only 1 of 5. Integration pages showed up only when the prompt named an integration: 7 citations across 4 answers, all of them integration prompts.
Our reading: ChatGPT answers each qualifier with the page that states it. A company-size question needs seat limits and plan tiers, which live on the pricing page. An integration question needs a page that names both products. A plain "best X" question has no qualifier to check, so the pricing page has less to add.
The GitHub prompt shows how literal this is. For the best project management tool that integrates with GitHub, all four sources ChatGPT cited were vendors' GitHub integration pages or setup docs. One vendor had both its integration page and its docs article cited. None was a blog post or a list.
So before you write anything new, find out which of your pages ChatGPT already cites for your buying prompts. Run your own qualified prompts and note which URLs appear as sources. If a competitor's pricing page keeps showing up and yours doesn't, that's the page to fix first.
Step 3: Build Pricing, Integration, and Help Pages ChatGPT Can Quote
Write each page so one paragraph answers one buyer qualifier. On pricing pages, state plan names, prices, seat or usage limits, and who each plan is for, in text. Give each major integration its own page naming both products and the job they do together. Keep setup articles public and crawlable, so ChatGPT can reach them.
Picture each buyer qualifier as a question the page must answer in one place. Team-size buyers want seat limits and the plan that fits. Someone on a budget wants a starting price and what it includes. A buyer comparing alternatives wants to know who you're for, and who you aren't. If that information sits behind a "contact sales" button, ChatGPT has nothing on your own site to quote. It may quote whoever does state it instead, such as a rival's comparison or a third-party list.
One outside study found a very different number. Position Digital's analysis of 278 unique prompts across six B2B SaaS categories, using July 2026 data, put pricing pages at 0.4% of ChatGPT's citations. Our figure was 30.1%. Its prompt set mixed informational, commercial, and branded questions, while every one of ours was a buying prompt.
We think buying stage explains most of the gap, though that's reasoning, not a test. Eleven of our 12 answers that cited a pricing page came from qualified prompts: size, price, or alternatives. The two samples differ in size and method, so neither number settles the question alone. If your buyers ask ChatGPT purchase questions, your pricing page is in play.
For integrations, a directory of logos gives ChatGPT little to quote. A page per major integration, saying what syncs and in which direction, gives it a sentence it can lift. Help articles work the same way. If your setup guides sit behind a login, ChatGPT can't cite them.
Add SoftwareApplication markup with offers, plus Organization markup. Treat it as hygiene rather than a citation lever. Our guide to what schema markup helps AI engines cite your pages covers the details.
Pricing pages raise their own questions, such as whether to publish prices at all. We cover them in our guide on whether your SaaS pricing page affects AI recommendations.
Step 4: Publish Fair "Best X" Lists That Perplexity Can Cite
Perplexity reads lists, including vendors' own. In our audit, a vendor's own "best X" or alternatives list appeared in all 25 of its answers. Those lists made up 30.1% of its citations. It cited no pricing pages and no help-center articles. The practical route to being a Perplexity source is a list page worth citing.
Of the vendors Perplexity named, 39.9% were cited through a list page on their own site. And 90.9% of its citations on vendor domains went to list pages rather than product pages. So on Perplexity, a vendor's own list often does the job that pricing and product pages do on ChatGPT.
Publish the lists honestly. Write category lists and "alternatives to X" posts that include real competitors, describe each one fairly, and say who each tool is for. A list that only praises you is a sales page, and we'd expect readers and engines to treat it that way. Rank yourself where you genuinely fit, not automatically first.
Alternatives lists matter most for challengers. On "alternatives to" prompts, the market leader's own site was cited in only 1 of 5 answers on each engine. Competitors' own "alternatives to" posts were cited on those prompts. Our guides cover getting into the "best tools" lists AI engines cite and whether comparison and alternatives pages help you get cited.
Step 5: What Gets You Named When Your Site Isn't Cited?
Off-site presence gets you named. ChatGPT cited the vendor's own site for only 41.1% of the vendors it named in our audit. Perplexity did so for 42.0%. For the rest, the answer cited other pages, or none, for that vendor. Independent lists were 29.1% of ChatGPT's citations and 50.0% of Perplexity's. Review sites appeared in 14 of 25 Perplexity answers, against 2 of 25 on ChatGPT.
This is where owned pages reach their limit. You can publish the clearest pricing page in your category and still miss an answer built from lists you're not in. The shortlist often comes from pages other people wrote. Your own site then supplies the details, if it's cited at all.
Review sites are the clearest engine split. They showed up in most Perplexity answers and almost no ChatGPT answers. So a G2 or Capterra profile likely matters more on Perplexity. Our guide on whether G2 and Capterra reviews decide AI recommendations goes deeper. Reddit barely registered on either engine: ChatGPT cited it in 0 of 25 answers, and Perplexity in 1 of 25.
The same Position Digital study points in this direction. It found that 66% of brand recommendations happened without ChatGPT citing that brand's website at all. It also found that the share of category listicles a brand appears in correlates with how often ChatGPT names it, at r = +0.64. That's a correlation, not proof that lists cause mentions.
Our reading: owned pages make you the cited source, and lists and reviews make you the named vendor. You need both, because the two engines lean on them differently. Contact the publishers of the independent lists in your category, and keep your review profiles current.
How Do You Know Your SaaS Is Getting Cited?
Run the same qualified prompts on each engine, logged out. Log two things separately: which of your pages are cited, and whether you're named at all. Being named and being cited came apart for most vendors in our audit, so one blended "AI visibility" score hides which layer changed.
A basic log covers it:
- Group prompts by qualifier. Write prompts for company size, integration, price, and alternatives, plus a plain "best X" for each category you sell into.
- Keep results per engine. Record ChatGPT and Perplexity in separate columns, since they share so few sources.
- Log cited and named apart. Note every URL of yours that's cited, and every answer that names you without citing you.
- Repeat monthly. One run is a snapshot. Wait for several before you trust a trend.
Our guide on how to measure AI share of voice for a SaaS product covers the method in more detail.
Frequently Asked Questions
How do I get my SaaS cited in ChatGPT?
Start by allowing OAI-SearchBot, which OpenAI uses to surface sites in ChatGPT search. Then make your pricing, integration, and help pages state the facts buyers ask about in plain text. In our audit, pricing pages were ChatGPT's most-cited page type, at 30.1% of its citations, mostly on size, price, and alternatives prompts.
Does Perplexity cite different pages than ChatGPT?
Yes, very different ones. For the same prompt, the domains the two engines cited overlapped by only 5.5% on average in our audit. ChatGPT leaned on vendor pricing, product, and help pages. Perplexity cited mostly lists, including vendors' own "best X" posts, and cited no pricing pages at all.
Should I block GPTBot?
That's a policy choice, and it's separate from search visibility. According to OpenAI's crawler documentation, GPTBot crawls content that may be used for training, while OAI-SearchBot surfaces sites in ChatGPT search. Sites that opt out of OAI-SearchBot won't be shown in ChatGPT search answers. You can block GPTBot and still allow OAI-SearchBot.
Does ChatGPT read SaaS pricing pages?
It cited them often in our audit. Pricing pages appeared in 12 of 25 ChatGPT answers, and 11 of those 12 came from size, price, or alternatives prompts. A Position Digital study found pricing pages were just 0.4% of ChatGPT's citations. Its prompts mixed informational, commercial, and branded questions, so we think buying stage explains most of the gap.
Why is my SaaS named but not cited?
Because answers often name vendors on the strength of other people's pages. In our audit, ChatGPT cited a named vendor's own site for only 41.1% of the vendors it named, and Perplexity for 42.0%. For the rest, the answer cited other pages, such as independent lists and review sites, or none for that vendor. That's why off-site presence matters.
Conclusion
Getting your SaaS cited in ChatGPT and Perplexity is a sequence, and the order matters. Let the search crawlers in first. Then build the pricing, integration, and help pages ChatGPT quotes, and the fair "best X" lists Perplexity reads. Finally, earn places on the independent lists and review sites that name you when your own site isn't cited. Measure each engine separately, or you won't know which step is working.
Agencies such as Kubnal Bridge specialize in running this per-engine audit and building the pages each engine reads. If you'd rather not run the prompts in-house, an outside audit is a sensible place to start.
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
- OpenAI, "Overview of OpenAI Crawlers," https://developers.openai.com/api/docs/bots
- Perplexity, "Perplexity Crawlers," https://docs.perplexity.ai/guides/bots
- Vercel, "The rise of the AI crawler," 2024-12-17, https://vercel.com/blog/the-rise-of-the-ai-crawler
- Position Digital, "Top ChatGPT Ranking Factors in B2B SaaS Niches," updated 2026-08-12, https://www.position.digital/blog/chatgpt-ranking-factors/
- Kubnal Bridge, SaaS citation audit: 25 software-buying prompts across ChatGPT, Perplexity, and Google AI Overviews (first-hand data)