Perplexity changes SEO because it answers questions directly and lists a handful of cited sources, so the goal shifts from ranking a blue link to being one of the pages it reads and cites. Pages that answer specific questions clearly, show expertise, stay crawlable and are referenced by other trusted sites have the best chance. Classic SEO foundations still apply, but the reward looks different.
Perplexity is an AI answer engine built by Perplexity AI. A user asks a question, the system searches the web, reads several sources and writes a summary with numbered citations. Users can ask follow-ups, switch to deeper research modes and filter sources. For many research-style queries, the user never visits a traditional results page.
How does Perplexity choose sources?
Perplexity does not publish a full ranking method. From observed behaviour and its own documentation, a few points are clear. It runs its own crawler, PerplexityBot, and also uses search results to find pages. It favours pages that directly match the question, and it tends to cite sources that are clearly structured and focused.
Because each answer cites only a few pages, competition is concentrated. Being the eighth result on Google still brings some clicks. Not being cited by Perplexity brings none from that answer.
Can Perplexity crawl your site?
Check this first. Some sites block AI crawlers in robots.txt without realising it, often through a security plugin or a CDN bot setting. Perplexity documents its user agents, including PerplexityBot, and states that it respects robots.txt for that crawler.
- Open your robots.txt and search for PerplexityBot, as well as wildcard rules that may block it.
- Check your CDN or firewall (Cloudflare, for example, has AI bot controls) for blanket blocks.
- Confirm important pages return a 200 status and render their main text without heavy client-side scripts.
Blocking is a legitimate choice for some publishers. It should be a decision, though, not an accident.
What content gets cited more often?
Pages that answer a question in their opening lines, then support the answer with specifics, do well in answer engines generally. Perplexity is no exception. Look at what it cites for your target questions and you will often see definitions, comparison tables, step lists and original data.
Practical content changes
- Put a direct two- or three-sentence answer under each question heading.
- Use descriptive headings that mirror how people ask questions.
- Add tables for comparisons and specifications; they are easy to quote.
- Show authorship, dates and sources so the page looks maintained.
- Publish original material: your own survey, benchmark or worked example gives others a reason to cite you.
Do backlinks and brand mentions still matter?
Yes, perhaps in a new way. Answer engines draw on pages that are already trusted and widely referenced. A brand mentioned in respected industry publications, review platforms and comparison articles has more chances to appear, either directly or through those third-party pages.
That makes digital PR and listings work part of AI visibility. If Perplexity keeps citing a particular directory or review site in your category, a complete and accurate profile there may be worth more than another blog post on your own domain.
How do you track Perplexity visibility?
Traffic from Perplexity appears in analytics as referral traffic from perplexity.ai. In GA4, check Traffic acquisition with session source, or build an exploration filtered for AI referrers. The numbers are often small, but visitors arriving from a cited answer tend to be well informed.
For citation tracking, keep a fixed list of questions and run them monthly:
- Ask each question in Perplexity and record whether your domain is cited.
- Note the position of the citation and the competing sources.
- Save the answer text when it mentions your brand, so you can check accuracy.
- Compare month to month rather than reacting to a single run.
What should you not change?
Do not rewrite every page into a list of questions, and do not chase AI visibility at the cost of the pages people already find through Google. Google still sends most organic traffic for most sites. The good news is that the changes answer engines reward, such as clear structure, real expertise and specific answers, also tend to help in Google.
Avoid publishing thin pages that target every possible question variant. Google’s scaled content abuse policy applies whatever the motivation, and Perplexity has little reason to cite pages that add nothing new.
Does structured data help with Perplexity?
Schema markup is not a confirmed Perplexity ranking factor, and no public statement says it is. It still has two benefits. It helps Google, which remains a source of discovery for many systems, and it forces you to state facts such as product names, prices, authors and dates in a clean form. Pages that express their facts clearly tend to be quoted accurately.
Use the types that genuinely fit the page, such as Article, Product, FAQPage or Organization, and validate them with Google’s Rich Results Test. Markup that does not match visible content is a risk, not a shortcut.