01Notes · 30 August 2026
Why Perplexity Cites Your Competitors Instead of You
Quick answer
Almost always the same cause: the source Perplexity drew from mentions them and does not mention you. The answer is assembled from other people’s pages, so being absent from those pages means being absent from the answer.
Perplexity lists its sources, which makes this the one platform where you can read the reason rather than guess at it. Four steps below.
Key takeaways
- Read the source list, not your own position. The domains cited alongside your competitor are the actual target.
- It is usually not about backlinks. 72.8% of AI-cited pages in 2026 research had no page-level referring domains.
- Ten prompts is enough to see the pattern. The same six to eight domains will keep recurring.
- Some fixes are immediate: a stale directory entry, an out-of-date comparison article, an inactive review profile.
- Perplexity moves fastest, so it is the best surface for testing whether a change landed.
- Do not optimise for Perplexity alone. Use it to see the source landscape the other platforms hide.
Step one: read the sources, not the ranking
The instinct when a competitor appears and you do not is to check your rankings. On Perplexity that is the wrong first move, because the answer was not assembled from a ranking. It was assembled from a handful of pages, and those pages are printed underneath.
Run the prompt. Then ignore whether you appear and write down every domain in the source list. That list is the thing you are actually competing for.
Step two: find the recurring domains
One prompt tells you very little, because these systems are non-deterministic. Ten prompts across your buyer questions will show you a pattern: the same six or eight domains recurring regardless of how the question is phrased.
In most categories that list is unsurprising once you see it — two or three review platforms, a comparison publisher, an industry title, a community thread, sometimes a competitor’s own documentation. What surprises people is how rarely their own site is on it.
| Source type | Why your competitor is there | How long it takes you |
|---|---|---|
| Review platform | An active profile with recent reviews | Weeks. A real request flow to real customers |
| Comparison article | The author included them and not you | Days, if the article can be updated. Ask |
| Directory or listing | A complete, current entry | Days. Often the fastest win on the list |
| Community thread | Someone recommended them, genuinely | Months. There is no shortcut and faking it is obvious |
| Industry publication | Earned editorial placement | Weeks to months, pitched not paid |
| Their own documentation | Specific, factual, structured, crawlable | Yours already. Check it is not blocked |
Step three: work out why they are in and you are not
Three explanations cover most cases, and they need different responses.
You are not mentioned in the source at all. The comparison article lists six vendors and you are not one of them. This is the most common and the most fixable, sometimes with an email to the author.
You are mentioned but described wrongly. The source has your pricing from two years ago, or lists a feature you removed. The model reads what is there. Correcting the source is faster than building a new one.
You are absent from the whole category conversation. Nothing authoritative connects you to the query family. This is the real gap and the slow one — months of earned placement, not a content sprint.
Worth saying plainly: it is usually not about links. In 2026 research, 72.8% of AI-cited pages had no page-level referring domains at all. Links were not the entry condition in that sample, which means a link-building programme aimed at this problem may be aimed at the wrong target.
Step four: separate the fast fixes from the slow ones
Once you have the source list and the reason, the work sorts itself into two piles. The fast pile is corrections: stale directory entries, outdated comparison articles, dormant review profiles. Days of work, and it usually moves first because Perplexity is retrieval-led and reflects changes quickly.
The slow pile is earning a place where you have none. That is pitched editorial work, and it takes months. Anyone promising it faster is describing something bought, which carries a different risk entirely.
The whole procedure is in the published method if you would rather run it yourself. It works.
What this means for you
- Do the ten-prompt source read before commissioning anything. It costs an hour and it changes what you commission.
- Check the fast pile first. A stale listing corrected this week beats a content plan starting next quarter.
- Use Perplexity to see, not to win. Its value is diagnostic long before it is commercial.
02Common questions
Questions this raises
Why does Perplexity cite my competitor and not me?
In most cases because the source it drew from mentions them and not you. Perplexity lists its sources, so this is one of the few AI visibility questions you can answer by reading rather than inferring.
Is it because they have more backlinks?
Often not. In 2026 research, 72.8% of AI-cited pages had no page-level referring domains at all. Links were not the entry condition in that sample.
How do I find which sources Perplexity uses for my category?
Run ten buyer prompts and record every domain listed, not just whether you appear. After ten prompts the same six or eight domains will keep recurring. That list is your actual target.
Can I get added to those sources?
Sometimes directly: a comparison article can be updated, a directory listing corrected, a review profile activated. Sometimes only by earning a place, which is slower. Both are off-page work, not on-site content.
How fast do changes show up in Perplexity?
Faster than anywhere else. It is retrieval-led, so a corrected fact or a new placement often appears within weeks. That makes it the best place to test whether something worked.
Should I optimise specifically for Perplexity?
Only if your buyers use it. Its audience is much smaller than Google's surfaces. What makes it valuable is diagnostic: it shows you the source landscape that other platforms hide.
Sources
Every figure on this page comes from one of the following. Where a claim is ours rather than a source’s, it carries an evidence label instead.
- Rankability — State of AI Search, 1,645 citation observations across 31 topics, 2026. Source for the 24.1% cross-platform overlap, the 73.5% ChatGPT exclusive citation rate and the first-100-words finding. Independent study
- Searlixa sampling across client engagements, labelled Observed with sample and date. Our own
Ask an assistant
Hand this page to an assistant and ask it to summarise or challenge the argument. If it gets something wrong about us, that is itself a useful finding.
→Next step