How to get your products cited in Perplexity.
Perplexity keeps citing review sites and competitors for queries you should own. Its shopping answers carry 3–7 cited sources each — more slots than any other engine — and its citation behavior differs enough from ChatGPT's that generic advice doesn't transfer.
Baseline your per-SKU citation rate on Perplexity, verify PerplexityBot access, then fix by failure mode: schema and crawl for MISSED, answer-coverage rewrites for BRAND-only, third-party grounding for low positions. eCommerce Insights tracks Perplexity citations per SKU on every refresh — see the Perplexity platform page.
The slow way: ten prompts, a private window, a spreadsheet
The manual baseline is worth doing once. Open Perplexity logged out in a private window. Write ten category-typical prompts the way buyers type them — three top-of-funnel ("best [category] for [use case]"), three feature comparisons, two competitor queries, one price-band, one scenario-specific. Run each, and record three things per prompt: whether your product was cited, the citation position (1–7), and which sources won the other slots. Keep the wording identical week to week; tiny rewordings change answers.
Ten prompts take 30–45 minutes a week, and the discipline decays exactly the way every recurring manual task does. Worse, single runs mislead: Perplexity's retrieval is non-deterministic, so a one-week swing of ±15 percentage points is noise, and you need three or more consistent weekly runs before a trend is real. Most teams quit before the data becomes meaningful.
The eCommerce Insights way
- Baseline automatically. Every SKU is tracked against category and intent prompts on Perplexity each refresh — CITED, BRAND, or MISSED per prompt, with position and the competing sources logged. History accumulates from day one.
- Verify PerplexityBot access. The agent-readability check confirms PerplexityBot is admitted in robots.txt and actually receiving 200s. A blocked bot is a hard zero no rewrite can fix — see fix robots.txt blocking AI crawlers.
- Fix by failure mode. MISSED on every prompt: crawl access and schema first. BRAND-only: the engine knows the brand but not the SKU — rewrite the first 300 characters of the PDP with quotable facts, add FAQ answer-coverage, complete the Product JSON-LD. Cited with wrong details: stale content; refresh and let it recrawl.
- Build third-party grounding for position. Perplexity weights independent corroboration heavily — review sites and category publications. Going from position 5+ to the top three is usually a grounding problem, not a PDP problem. The observed patterns are documented in Perplexity shopping citation patterns.
- Track weekly, hold wording stable. The dashboard plots citation rate and average position per SKU; prompt runs store every verbatim answer so you can read exactly what changed when a slot flips.
What "good" looks like
Directional reads for a mid-market D2C SKU, from eCommerce Insights tracking as of early 2026 (illustrative):
Also watch position: the first three citation slots carry visible prominence in Perplexity's UI, and inferred click weight concentrates there. A SKU stable at position 6 has different work to do than a SKU missing entirely.
Ask AI about this job
Have your favorite AI engine apply this walkthrough to your category.
Frequently asked questions
How does Perplexity decide which products to cite?
Why is Perplexity different from ChatGPT for this job?
Does citation position matter, or just being cited?
What about Perplexity Shopping and Buy with Pro?
How long until PDP changes show in Perplexity citations?
Seven slots per answer. Claim yours.
Per-SKU Perplexity tracking with the fix attached to every miss.