ChatGPT searches the live web through Bing. Copilot runs on Bing. Google AI Overviews read Google's index. Perplexity is the outlier. It runs its own crawler, PerplexityBot, builds its own search index, and retrieves pages in real time when a user asks a question. That means the signals that get a contractor into ChatGPT or Copilot do not guarantee the same contractor appears in Perplexity. A trades business can have a perfect Google Business Profile, a Bing Places listing, and 200 Google reviews, and still be invisible to Perplexity if PerplexityBot has never read the website.
Perplexity's own documentation is clear about this. PerplexityBot is designed to surface and link websites in Perplexity's search results. It is not used to crawl content for AI foundation models. It reads pages, indexes them, and cites them when a user asks a question that the page answers. (Perplexity Crawlers Documentation)
For a plumber, HVAC contractor, or electrician who has been told to "optimize for AI search," this is the catch: there is no single AI search. Each tool reads different sources, and Perplexity reads only what its own crawler can reach.
How PerplexityBot works
PerplexityBot visits websites, reads the content, and adds pages to Perplexity's search index. When a user asks Perplexity "who is a good plumber near me in Bergen County," Perplexity does not send that query to Google or Bing. It searches its own index, retrieves the most relevant pages, reads them, and writes an answer with inline citations linking back to the source pages.
This is different from ChatGPT, which uses outside search partners for live web results. ChatGPT can pull local business data from third-party providers and display it in a data card beside the answer. Perplexity does not have that same partnership layer for local business data in the same way. It relies more heavily on what its crawler has read from the open web. (Pluspoint.io, 2026)
The practical difference for a contractor is this: if PerplexityBot has not crawled the contractor's website, or if the website blocks PerplexityBot in its robots.txt file, Perplexity has no source to cite. The contractor does not appear in the answer. A competitor whose website PerplexityBot can read gets the citation instead.
The robots.txt problem
Many contractor websites are built on platforms that add default robots.txt rules blocking unfamiliar crawlers. Some security plugins and firewalls block any bot that is not Googlebot or Bingbot. A contractor who launched a website five years ago and never checked the robots.txt file may be blocking PerplexityBot without knowing it.
Checking takes about 30 seconds. Visit the website's domain followed by /robots.txt. Look for a line that says "User-agent: PerplexityBot" with a "Disallow" rule beneath it. If that line exists, PerplexityBot is being told not to crawl the site. The fix is to either remove the PerplexityBot block or add a rule that explicitly allows it.
Perplexity's documentation recommends allowing PerplexityBot in robots.txt and permitting requests from their published IP ranges. The full user-agent string is: Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; PerplexityBot/1.0; +https://perplexity.ai/perplexitybot)
A contractor who wants to check this quickly can run the free AI visibility audit at firstcallvisibility.com, which checks whether PerplexityBot and other AI crawlers can access the site.
What PerplexityBot reads vs what it ignores
PerplexityBot reads the text content of web pages. It does not read JavaScript-rendered content the same way a browser does. If a contractor's website loads its service descriptions, phone number, and service area through JavaScript after the initial page load, PerplexityBot may see a blank page. Static HTML content gets crawled. JavaScript-rendered content may not.
This matters for trades businesses because many modern website builders render content through JavaScript frameworks. A plumber's website might look great in a browser but present almost nothing to a crawler. The fix is server-side rendering or static HTML fallbacks for key content like service pages, contact information, and service areas.
PerplexityBot also reads structured data. A website with LocalBusiness schema markup gives the crawler a clean, machine-readable record of the business name, address, phone number, services, and service area. A website without schema forces the crawler to guess which text on the page is the business name, which is the address, and which paragraph describes the services. Schema makes the crawl cleaner and the citation more accurate.
How to test Perplexity visibility
The test is the same one that works for every AI tool, with one adjustment for Perplexity's citation model. Ask Perplexity what a homeowner would ask:
- "Who is a good plumber in [county], [state]?"
- "What plumbers near [city] do emergency repairs?"
- "Best HVAC contractor in [county] for AC replacement?"
Read the answer. Write down which businesses Perplexity names and which sources it cites. The sources matter because they show what PerplexityBot has crawled and indexed. If Perplexity cites a competitor's website but not the contractor's, the competitor's site is being crawled and the contractor's is not.
Run the same questions in ChatGPT, Gemini, and Copilot. Compare which businesses each tool names. The differences reveal which sources each tool reads and where the gaps are. A contractor who appears in ChatGPT but not in Perplexity has a PerplexityBot problem, not a general AI visibility problem.
What to do this week
- Visit the website's /robots.txt file. Confirm PerplexityBot is not blocked.
- Confirm the website renders service pages as static HTML, not JavaScript-only content. If unsure, view the page source in a browser and search for the service description text. If it does not appear in the source, it is JavaScript-rendered and may not be crawled.
- Add LocalBusiness schema markup to the homepage and each service page. Include the business name, address, phone, services, and service area.
- Give each major service its own page with specific, crawlable content. "AC repair in Bergen County NJ" with 200 words of real text outperforms "HVAC services" with a contact form and nothing else.
- Run the free AI visibility audit to see which crawlers can access the site and which signals are missing.
- Ask Perplexity the same five questions a homeowner would ask, in a fresh session, and note which businesses appear and which sources get cited.
Why Perplexity matters even if it is smaller than ChatGPT
Perplexity had roughly 78 million monthly active users as of mid-2026, per reporting from The Information. That is a fraction of ChatGPT's user base. But Perplexity users behave differently. They are researching, comparing, and clicking through to cited sources. A Perplexity answer that cites a contractor's website includes a direct link. That link gets clicked by a homeowner who is already evaluating providers, not someone just asking a casual question.
The conversion intent of a Perplexity click tends to be higher than a ChatGPT click because Perplexity's interface is built around source links. A contractor who appears in a Perplexity answer with a link to a service-specific page is getting traffic from someone who has read the answer, seen the source, and decided to visit.
FAQ
Does Perplexity use Google or Bing for local business recommendations?
No. Perplexity uses its own web crawler, PerplexityBot, to build its search index. PerplexityBot reads websites and surfaces them as cited sources in Perplexity's answers. Perplexity does not pull from Google Search, Google Maps, or Bing Places the way ChatGPT and Copilot do. A contractor who has optimized their Google Business Profile but blocked PerplexityBot in robots.txt will not appear in Perplexity answers.
How do I check if PerplexityBot can read my contractor website?
Open the website's robots.txt file by visiting yourdomain.com/robots.txt. Look for a line that says "Disallow" under "User-agent: PerplexityBot". If PerplexityBot is disallowed or if the entire site is blocked with "Disallow: /", Perplexity cannot crawl the site. To fix it, either remove the PerplexityBot block or add a specific allow rule. Perplexity's own documentation recommends allowing PerplexityBot in robots.txt and permitting requests from their published IP ranges.
What is the difference between ChatGPT and Perplexity for local contractor recommendations?
ChatGPT searches the live web through outside search partners, including Bing, and can use local listings from third-party providers. Perplexity uses its own crawler, PerplexityBot, to build an independent search index and retrieves pages in real time. ChatGPT may name businesses from Yelp-sourced data cards. Perplexity names businesses based on what its crawler has read and indexed. A contractor can appear in ChatGPT but not in Perplexity, or vice versa, because the two tools read different sources.
How many homeowners use Perplexity for local contractor recommendations?
Perplexity had roughly 78 million monthly active users as of mid-2026, per reporting from The Information. That is smaller than ChatGPT's user base, but the users who go to Perplexity tend to be researching purchases, comparing options, and reading cited sources. Those are homeowners further along in the decision process. A contractor who appears in a Perplexity answer with a direct link to their website gets a click from someone already evaluating providers.
What should a trades business do first to get visible on Perplexity?
Check that PerplexityBot is not blocked in the website's robots.txt file. Then ensure each service has its own page with specific, crawlable content. Add LocalBusiness schema markup so PerplexityBot can parse the business name, address, phone, and service area. Ask ChatGPT, Perplexity, and Gemini what a homeowner would ask about the trade in the service area, and compare which businesses each tool names.