In the Italian food sector, six sites account for 60% of all citations on Perplexity — real data across a hundred analyzed answers. If you're not in that narrow group, you get the crumbs: a few sporadic mentions that build nothing. Meanwhile, those six domains get recommended every day to thousands of users searching for exactly what you sell. Breaking into the core is hard but not impossible — and there's a precise path to do it.
Perplexity weighs sources with a proprietary scoring system. I analyzed 100 answers: 6 domains dominate 60% of the citations in the Italian food sector. The rest split the crumbs.
This is the figure that changes how you have to approach Perplexity. It’s not a lottery, it’s not a random algorithm, it’s not “whoever publishes the most wins.” It’s a weighting system that rewards a narrow core of domains and leaves everyone else fighting over the remaining 40%. If you’re a dairy in Lodi producing Grana Padano DOP, you need to know whether you’re in the core or among the crumbs — because this determines whether Perplexity sends traffic to your site or to the site of a consortium selling the very same product.
In my previous articles I explained how ChatGPT, Claude and Gemini work with sources. Perplexity plays in a league of its own: it’s the only AI that puts a clickable link next to every statement, and this turns the citation into measurable traffic. I’ll explain it with the data in hand.
What does it mean to “be cited” on Perplexity for your brand?
Perplexity behaves differently from other AI engines. For example, for a query like “best Grana Padano DOP aged 24 months,” you don’t get a long block of text with generic sources at the bottom. You get numbered paragraphs, each with its own clickable link to the source. Click it, and you land on the original site.
This technical difference has an enormous commercial consequence. On ChatGPT you can be “inside the answer” without anyone ever reaching your site — the model synthesizes, the user reads, end of story. On Perplexity every citation is an open door to your domain.
From this follows a practical rule: the value of a Perplexity citation is measurable in euros of traffic, while the value of a ChatGPT citation is measurable in brand awareness (harder to quantify). If you run a dairy in Lombardy and sell Grana Padano DOP online or through third-party e-commerce, Perplexity is the first AI to focus on. I’m saying this with data, not hearsay.
Why Perplexity’s scoring is different from the rest
In previous articles on E-E-A-T for AI and backlinks as a citation proxy I explained how AI engines assess authority. Perplexity takes those principles and applies them more strictly: instead of averaging across many sources, it picks a narrow core and keeps coming back to it repeatedly.
The mechanism isn’t officially disclosed, but the pattern is clear to anyone who observes it. Perplexity seems to use a mix of three signals: domain authority (backlinks, brand recognition, presence in the knowledge graph), content freshness (it prefers pages updated within the last 12 months) and information density (pages with structured data, tables, and FAQs get cited more often than narrative articles).
The operational consequence is that optimizing for Perplexity is not optimizing for Google. A page that ranks in the top 3 on Google may never appear on Perplexity, and vice versa. It’s a separate channel that requires a separate strategy.
Many Lombard dairies I follow optimize their Grana Padano product page with classic keyword stuffing.
The observation I made: 100 Italian food answers
Let me share the figure I previewed at the start, with a transparent method. For two months I monitored Perplexity’s answers to 100 queries about the Italian food sector: DOP cheeses, wines, oils, artisanal pasta, preserves. Queries in Italian, geolocation set to Italy, neutral account.
Result: 6 domains collected 60% of all citations. Wikipedia, the websites of the DOP protection consortia, two major Italian food publications, a specialized e-commerce portal and an institutional site. The remaining 40% was split among roughly 80 different domains, each with a few scattered citations.
Limits of the test: a sample of 100 queries, two months of observation, a single language. It’s not a scientific study, it’s a longitudinal observation that confirmed a pattern I suspected. For your specific sector you might see different distributions, but the principle “a few domains dominate, many collect crumbs” repeats in every vertical I’ve monitored.
For a dairy in Lodi producing Grana Padano DOP, this means one precise thing: if you’re not among the 6 dominant domains, you’re competing with another 80 sites for the residual 40% of citations. Your statistical odds are low, but not zero — and that’s exactly where the game is played.
Use Google’s Rich Results Test to verify whether your product pages have Product, Organization, and FAQPage schema markup
The test you can run in 20 minutes
Open Perplexity in incognito (to avoid personalization) and run 10 queries relevant to your sector. For the Lombard dairy, try queries like: “best long-aged Grana Padano DOP,” “difference between Grana Padano and Parmigiano Reggiano,” “where to buy Grana Padano DOP online,” “Grana Padano DOP 30 months characteristics.”
For each answer, note on a sheet: which domains appear in the numbered citations, in what position (1, 2, 3…), and whether they appear more than once in the same answer. After 10 queries you’ll have a rough but useful map of the “6 dominant domains” in your sector.
Then open Google Analytics 4 and filter traffic by the “perplexity.ai” referrer. If the number is zero or close to it, you’re among the crumbs. If you see regular sessions, you’re at least on the periphery of the core. The GA4 figure is the most reliable truth you have, because it measures the real result, not an opinion about how visible you are.
Binary decision threshold: if your monthly sessions from Perplexity are under 20, you need to act. If they’re above 100, you have an asset to protect and strengthen. Between 20 and 100, you’re in the growth zone — and it’s the right moment to invest.
Here are the most frequent mistakes
Treating Perplexity like Google. Many Lombard dairies I follow optimize their Grana Padano product page with classic keyword stuffing. Perplexity ignores that kind of optimization. It wants pages with structured data, nutritional tables, FAQs, and short, dense paragraphs.
Not monitoring the perplexity.ai referrer. Without GA4 data, you don’t know whether you’re improving or getting worse. I’ve seen companies invest in content “for the AI” without ever checking whether the AI sent traffic back. The referrer is your only reliable KPI.
Confusing citation and link. Being named in a Perplexity answer without being among the numbered sources is worth nothing in terms of traffic. It only counts as a brand mention, which is a long-term asset but doesn’t bring sessions in the short term.
Ignoring the competitors cited in your place. If the query “Grana Padano DOP 24 months” cites the consortium and three e-commerce sites but not you, you need to figure out why. Those 4 domains have something you don’t: probably complete schema markup, structured reviews, presence in Google’s knowledge graph.
What you can concretely do this week
- Open Perplexity, run 10 queries for your sector, map the 6 dominant domains
- Open GA4, filter traffic by the “perplexity.ai” referrer for the last 90 days — note the number
- Identify the 3 pages on your site that get the most Perplexity citations (if any): they’re your asset, protect them
- Compare the structure of your pages with that of the 6 dominant domains: look for tables, FAQs, and structured data you don’t have
- Use Google’s Rich Results Test to verify whether your product pages have Product, Organization, and FAQPage schema markup
This is an entry-level audit. The real analysis — the one that tells you why Perplexity chooses one domain over another in your sector — requires professional tools and months of observation. But these 5 steps give you an honest baseline to start from.
