7 minute read
ChatGPT vs Perplexity: how to show up in both
How ChatGPT and Perplexity search, crawl and cite, which sources each favors, what robots.txt means, and how to show up and measure it.
ChatGPT and Perplexity both answer questions with a short list of businesses and a handful of citations. They do not work the same way, they do not lean on the same sources, and a robots.txt line that helps you in one can quietly remove you from the other. Here is what each company has published about how it works, and what that means for a business that wants to be named.
How each one searches
ChatGPT. OpenAI says ChatGPT search uses third-party search providers plus content supplied directly by partners, with data providers for categories such as weather, stocks, sports, news and maps. Answers show sources in a sidebar. OpenAI says any public website can appear in ChatGPT search, and that the key requirement is not blocking its search crawler.
Perplexity. Perplexity describes itself as searching the internet in real time and citing what it uses with numbered footnotes linking to the original pages. Its Research mode runs dozens of searches and reviews hundreds of sources. Perplexity also runs its own crawler to surface and link sites in its results.
Neither company publishes a ranking formula. What they do publish is how their crawlers behave, which is where most fixable problems are.
The crawlers, side by side
| User agent | What it does | Robots.txt |
|---|---|---|
| OAI-SearchBot (OpenAI) | Surfaces sites in ChatGPT search results | Respected. Blocked sites do not appear in search answers, though they may still show as a navigational link |
| GPTBot (OpenAI) | Collects content that may be used to train OpenAI's foundation models | Respected. Blocking it opts out of training only |
| ChatGPT-User (OpenAI) | Fetches a page when a user's action in ChatGPT or a custom GPT requires it | May not apply; not an automatic crawler and does not decide search inclusion |
| PerplexityBot | Surfaces and links sites in Perplexity results; Perplexity says it is not used to train foundation models | Respected |
| Perplexity-User | Fetches a page when a user's question needs it, and may link to it | Generally ignored, because a user triggered it |
OpenAI states each setting is independent: you can allow OAI-SearchBot and block GPTBot. Both companies say robots.txt changes can take about 24 hours to take effect.
Robots.txt: the most common self-inflicted problem
We regularly see sites that blocked "AI bots" as a group in 2024 and are now invisible in AI search. If you want to be recommended, allow OAI-SearchBot and PerplexityBot. Whether you allow GPTBot is a separate business decision about training, and blocking it does not remove you from ChatGPT search.
Robots.txt is only half of it. Firewalls and bot protection often block these crawlers even when robots.txt allows them. Perplexity's documentation recommends allowing its published IP ranges and matching both user agent and IP in rules on services like Cloudflare and AWS WAF. OpenAI also publishes IP ranges for its crawlers. Check your server or CDN logs to confirm the bots actually get a 200 response.
What each tends to cite
A March 2026 analysis of 30 million US sources by Peec found different favorites. ChatGPT's top cited domains were Wikipedia, Reddit, Forbes, TechRadar and LinkedIn. Perplexity's were Reddit, YouTube, LinkedIn, Wikipedia and G2. In practice:
- ChatGPT leans toward editorial and reference sources: encyclopedic pages, major publications, review sites.
- Perplexity leans further into community and video content, and G2 shows up in its top five, which matters for software companies.
- Both cite Reddit, LinkedIn and Wikipedia heavily.
The same analysis notes Facebook blocks Perplexity's bots through robots.txt, so content that lives only on Facebook is unlikely to be cited there.
Local versus national: two different games
Multi-location and local businesses. For "near me" style questions, OpenAI lists maps among its data partner categories, so your business profile data and listings feed the answer. Consistent name, address, phone, hours and categories across listings, strong recent reviews and clear location pages with LocalBusiness schema are the levers. Inconsistent data across directories is the most common reason a location gets skipped.
B2B and software companies. Here the question is usually "best X for Y." Answers lean on comparison articles, review platforms such as G2, Reddit threads, LinkedIn and analyst or trade coverage. Your own site needs clear pages on who you serve, pricing or price ranges, integrations and honest comparisons, because those are the facts assistants quote.
What to fix, platform by platform
| Fix | ChatGPT | Perplexity |
|---|---|---|
| Allow the search crawler | OAI-SearchBot in robots.txt and firewall | PerplexityBot in robots.txt and firewall, with IP allowlisting |
| Third-party presence | Editorial coverage, Wikipedia accuracy, review sites | Reddit, YouTube, LinkedIn, G2 for software |
| Local data | Listings and maps data consistent everywhere | Listings plus pages it can fetch and cite directly |
| Your own pages | Specific, quotable answers and schema | Same; fresh, well-structured pages get cited |
How to measure it
- Referral traffic. OpenAI says ChatGPT adds utm_source=chatgpt.com to referral links. Track it in your analytics, along with referrals from perplexity.ai.
- Crawler logs. Count hits and response codes for OAI-SearchBot, PerplexityBot, ChatGPT-User and Perplexity-User. A spike in 403s means a firewall problem.
- A fixed prompt set. Write 20 to 50 questions buyers actually ask, by location or by category, and run them monthly in both tools. Record whether you are named, in what position, and which sources are cited.
- Citation sources. When a competitor is named and you are not, look at what was cited. That is your to-do list.
Answers vary from run to run, so look at trends across many prompts, not a single screenshot.
Which should you prioritize?
Most businesses should not choose. The technical fixes, allowing the right crawlers and cleaning up your data, help both at once. If you have to sequence the rest: local and multi-location businesses usually get more from listings accuracy and reviews, which help everywhere. B2B and software companies should pay close attention to Perplexity's lean toward Reddit, YouTube and G2, and to ChatGPT's lean toward editorial coverage and reference pages.
How we can help
Our AI Visibility Audit ($2,500) checks your crawler access, runs a prompt set across ChatGPT and Perplexity and gives you a prioritized fix list. See how this works for multi-location businesses and for B2B and online businesses, try the free checks in our tools hub, or book a 60-minute expert call for $450.