Why Perplexity isn't citing your B2B case studies

The research engine, not the search engine

When someone asks Perplexity 'what results do companies see from implementing X,' it isn't ranking pages the way Google would. It's trying to assemble an answer out of specific, verifiable claims it can attribute to a source. A case study that reads like a magazine feature — quotes, adjectives, a narrative arc — gives it almost nothing to extract.

This is why a company can have a dozen glowing case studies and still get zero mentions when a prospect asks an AI assistant to compare vendors. The content exists, but it isn't shaped like something a model can lift a fact from. Your competitor with three plainer case studies and one clear number in each one wins the citation.

Ungating for crawlers, and the trade-off that comes with it

A gated PDF behind a lead form is invisible to Perplexity's crawler, ChatGPT's browsing mode, and Claude's web tool. None of them fill out forms. If your best proof points live behind gates, they don't exist for AI visibility purposes, full stop.

The trade-off is real: gated content captures leads, and marketing teams are right to be nervous about giving that up. The fix isn't to ungate everything. Publish a full, ungated HTML version of each case study with the real numbers in it, and keep a longer gated asset (extended interview, ROI calculator, editable template) as the lead magnet. You get the citation and still have something worth trading for an email address.

Structure the case study synopsis-first

1. 01Open with a 2-3 sentence synopsis stating the company, the problem, and the headline result — before any background or narrative.
2. 02Follow with a short 'at a glance' block: industry, company size, timeframe, and the primary metric that moved.
3. 03Move the origin story and process detail below that, for readers who want context.
4. 04Close with the specific numbers again in a table, plus a quote attributed to a named person with a title.
5. 05Add an FAQ block addressing the two or three questions prospects actually ask about that use case ('how long did implementation take,' 'what did it replace').

Numbers beat adjectives

"Significant improvement in efficiency" is unusable to a model trying to answer "how much time does X save." "Cut manual invoice processing from 6 hours a week to 40 minutes" is a sentence Perplexity can quote directly and attribute to you by name.

Audit your existing case studies for every adjective doing the work a number should be doing — 'dramatically,' 'substantially,' 'game-changing' — and replace each one with the underlying figure, even an approximate one with a caveat. A hedged real number beats a confident vague claim every time a model is deciding what to cite.

Schema markup for professional services and results

Case studies benefit from a mix of Article schema and, where it applies, Review or ClaimReview-style structured markup around the specific result. At minimum, mark up the client organization, the date, and the author. If you publish quantified outcomes regularly, consider a Dataset schema entry for the underlying numbers so they're addressable independently of the surrounding prose.

This isn't about gaming a rich snippet. It's about giving the crawler an unambiguous, machine-readable version of the same fact that's sitting in your paragraph, as a backup path if the parsing of the prose fails.

Monitoring your Perplexity share of voice

Run your own product and your top two or three competitors through Perplexity monthly with the exact questions a prospect would ask ('best X for mid-size healthcare companies,' 'X vs Y pricing'). Log which sources it cites for each answer.

If your case studies never show up as sources on questions where they're directly relevant, that's a structural problem with the content, not a fluke. Tools like Mangools' AI Search Watcher and Frase can automate part of this tracking so you're not doing it manually every month.