You are testing your visibility on Perplexity using the free version — but your ideal client, the buyer who makes the purchasing decisions, almost certainly uses the paid version. The problem is concrete: the two versions, on the same question, can cite different sources, because the Pro one digs deeper and weighs the authority of sources differently. You could be visible to those who don't buy and invisible to those who sign the contracts. There are quick tests that reveal exactly where you are positioned on each version — and how to course-correct.
The users who choose to pay for AI search are, by definition, highly qualified: engineers, doctors, lawyers, entrepreneurs. Whoever decides to spend 20 dollars a month on a Pro subscription doesn’t do it out of boredom. They do it because they are looking for better answers, on more advanced models, with more sources retrieved and a level of precision that the free version cannot guarantee. If you appear only in the answers of the base version, you risk never showing up in front of that 1% who makes the important purchasing decisions.
And here is the problem that few people see: Perplexity Pro and Perplexity Free, on the exact same query, can cite different sources. If your ideal client uses the Pro version and you have optimized with the Free one in mind, you are talking to the wrong person.
Let me explain how I verified the pattern, what you can test in twenty minutes in your own sector, and why, to show up in AI answers, it pays to understand which Perplexity you are talking to.
What really changes between Free and Pro
Perplexity Free uses a base model and performs retrieval on a limited number of sources per query (“doing retrieval” means searching, selecting and pulling from the web the most relevant sources to answer a query).
Perplexity Pro lets you choose advanced models (GPT-4 class, Claude class, Sonar Large), activates Pro Search mode which performs deeper retrieval with more rounds of search, and includes features like Spaces, file upload and dedicated academic searches.
The point that matters to us is not technical, it is about positioning: the Pro version pulls from more sources, weighs authority differently and goes deeper before synthesizing the answer. A query like “best destination wedding masserie in Puglia” on Free may stop at the first 5-7 aggregator sources (TripAdvisor, common wedding blogs, the odd Vogue Sposa article). The same query on Pro Search can reach 15-20 sources, including reviews from international wedding planners, destination photographers’ blogs and the websites of the masserie themselves, if they have enough authority.
I already touched on this distinction in previous articles on how the backlink works as a citation proxy and on the implicit reference weight: more “retrieved” sources means that the authority signal counts more than the signal of aggregated popularity.
Why one version is worth more than the other for your target
Think about who searches for “luxury masseria for a wedding in Puglia”. It’s not the twenty-six-year-old organizing a bachelor party. It’s an American or German couple, aged 32-45, with a budget of 80-150k for the wedding, who is weighing Italy vs Provence vs Mykonos. They probably already have a wedding planner and almost certainly pay for a Pro subscription to AI search in order to get detailed reports on the venues before visiting them.
That target, on Perplexity, almost always moves on the Pro version. The same goes for the B2B buyer searching for “best supplier of special steels for aerospace Italy”: whoever pays for the subscription is looking for answers that justify important business decisions.
Translated into a benefit for you: if your masseria in Ostuni shows up in Free but not in Pro, you are intercepting the weekend tourist, not the high-budget bride or groom. To show up in the AI answers that really count for your bottom line, you need to optimize for Pro behavior.
Testing only the Free version because “it’s the same anyway”.
The reverse engineering I did: Perplexity Pro vs Perplexity Free
To understand how much the sources change between Pro search and the free one, I took a real case: a luxury masseria in Ostuni (province of Brindisi) that organizes destination weddings, with a primary target of US/UK/DACH couples. I built 10 queries that an international wedding planner would actually make, and I ran them on Perplexity in both Free mode and Pro mode (with Pro Search active).
The queries (in English, as an international user would phrase them):
- “best luxury masserie wedding venue Puglia”
- “destination wedding Ostuni Italy reviews”
- “private masseria rental wedding 100 guests”
- “Puglia wedding photographer recommended venues”
- “olive grove wedding venue south Italy”
- “best masseria for symbolic ceremony Brindisi”
- “wedding planner Puglia trusted venues”
- “luxury accommodation wedding party Ostuni”
- “exclusive use masseria Italy 3 days wedding”
- “Puglia destination wedding cost guide”
What came out, on a small but consistent sample:
- Perplexity Pro cited on average 12 sources per query, while the Free version stopped at 5.2 sources. Pro search runs more scanning passes and gathers far more information.
- Less than 30% of the sources were shared between the two versions. Most of the sites cited in the Pro answer did not appear at all in the Free answer.
- The Pro version drew much more often from specialized niche sources: international wedding planner sites (Junebug Weddings, Style Me Pretty, Destination I Do) and blogs of destination photographers with high sector authority.
- The Free version showed a strong reliance on large generalist aggregators (TripAdvisor, Booking, local tourism portals) and on the venue’s official website, if well optimized.
- The same masseria was cited on Perplexity Pro in 7 queries out of 10, while on Perplexity Free it appeared in only 3 queries out of 10. The same engine, on the same question, produced two very different visibility results.
Limitations of the test, stated up front: it’s 10 queries in a single sector, in English, with a single reference brand. It’s an indicative test, not a sector study. But the pattern is so clear-cut that it’s worth replicating on your business before taking for granted that your brand is visible. The real analysis, on stable statistical volumes, requires professional tools and session logs distributed over several months.
The operational conclusion, however, is clear: base search and advanced search return almost disjoint worlds. If you are aiming at a high-spending target, you cannot base your tests only on the free version.
Note the 3-5 recurring domains on Pro that never show up on Free: those are your citation source to win over.
The test you can run in twenty minutes in your own sector
You don’t need to have a Pro account to start. The twenty minutes of the free trial are enough, or a client or colleague who lends you their login for a few queries.
Procedure:
- Choose 5 real queries that your ideal client would make. Not the classic SEO keywords, but the conversational questions. For the Ostuni masseria: “best masseria for a wedding in Puglia 100 guests”, “masseria with a centuries-old olive grove for a symbolic ceremony Brindisi”, and so on.
- Run each one on Perplexity Free, in incognito (so you don’t pollute the results with your history).
- Run the same 5 on Perplexity Pro with Pro Search active, choosing the most advanced model available (Sonar Large or Claude class).
- For each query note: number of cited sources, which domains, in what order, and whether your brand appears.
Decision thresholds:
- If you appear in at least 3 queries out of 5 in both versions: you are well positioned, optimize at the margin.
- If you appear only in Free: you probably have generalist popularity signals (TripAdvisor, Google reviews) but little niche authority. To intercept the Pro target you need to work on mentions in specialized sector sources.
- If you appear only in Pro: a rare but fortunate situation, you have niche authority. The popular signal is missing, work on reviews and UGC (User Generated Content).
- If you never appear: you are not in the AI answers for that target. Full stop.
The mistakes I see more and more often
Testing only the Free version because “it’s the same anyway”. It’s not true. In the tests I’ve run across different sectors (food B2C, hospitality, professional firms) the delta of cited sources between Free and Pro ranges from 30 to 60%. Considering the behavior of the two models identical is an evaluation error.
Optimizing only for ChatGPT thinking you’ll also cover Perplexity. They are engines with different retrieval logics. A page that works on ChatGPT with browsing may not work on Perplexity Pro because the latter weighs implicit citations and source freshness differently.
Ignoring the language of the target. A masseria that optimizes only in Italian and then wants to intercept American wedding planners on Perplexity Pro starts at a disadvantage: the Pro model predominantly pulls sources in the language in which the query is written.
Thinking that citations on large aggregators are enough. Portals like TripAdvisor and Booking move the Free version. For the Pro version you need mentions in specialized sector sources: blogs of industry photographers, niche magazines, international sector directories.
What to do concretely this week
- Open Perplexity Pro (even just on trial) and the Free version. Run your 5 target queries in incognito on both.
- Compare the cited sources. Note the 3-5 recurring domains on Pro that never show up on Free: those are your citation source to win over.
- Look for editorial contacts in those 3-5 domains. A guest contribution, a case study or a mention in a sector roundup is worth more than ten generalist backlinks.
- Check that your official page is readable by AI engines: use Google’s Rich Results Test to check the Organization and Event schema, and a tiktokenizer to see how your brand description is tokenized.
- Compare with the 3-5 competitors that Perplexity Pro cites in your sector. If they have mentions on niche publications and you don’t, you know where to go knocking.
Where does this discussion lead?
Understanding the delta between Perplexity Pro and Free is one piece of the larger work: covering every AI engine with awareness of who is really on the other side. The principle of “being in AI answers” is not an abstract objective, it is a geography of different targets on different engines, and each engine has sub-behaviors that change with the subscription tier.
In the next articles of this series on AI platforms I’ll explain how the models behave when they activate reasoning mode, why Claude in the Anthropic Console cites different sources compared to Claude.ai consumer, and what changes between free Gemini and Gemini Advanced on the same commercial queries.
For now, start with this: there is no such thing as “being on Perplexity”. There is being on which Perplexity, for which user, at which moment.
