ProlinkageGet your free report
Home/Resources/AI Shortlist Plays
AI Shortlist Plays / Proof engines can quote

How do we write case studies that AI engines can quote?

Open every case study with one sentence an engine can lift whole: who, what changed, from what to what, over which dates, measured where. Keep it as text on its own page, with the evidence beside it.

By ProlinkagePublished 8 November 2026Play 21 of 40 · Edition 1

Effort: 4 to 6 hours per case study

Who does it: Marketing lead, with the account owner

Best for: B2B firms of 20 to 500 people with at least one result they can evidence

Evidence: Moderate (Aggarwal et al., KDD 2024; our test, 6 November 2026)

01 / WHY IT MATTERS

Why it matters.

When ChatGPT explained its shortlist for our own category on 6 November 2026, it justified each pick with published case studies and named B2B clients. Proof that can be quoted is what the engine used to decide.

Research points the same way. The GEO study by Aggarwal and colleagues (accepted at KDD 2024; arXiv version of 28 June 2024) tested rewrites of web pages and found that adding citations, quotations and statistics could “boost visibility by up to 40%” in generative engine responses. A case study that states a result with its unit, period and source gives the engine exactly that material.

02 / SELF-CHECK

The 10-minute self-check.

Open your three strongest case studies and score each one, one point per yes:

  • The first paragraph states the result with a number, a unit and a date range.
  • The data source is named (GA4, Search Console, your CRM).
  • The client is named, or described by industry, country and size.
  • The numbers are in the page text, not only inside an image or a PDF.

Below 9 out of 12, rewrite the openings before you write anything new.

03 / STEPS

Steps.

  1. Choose results you can back with a dated screenshot. Leave out any you cannot.
  2. Write the opening as one sentence: “[Client] grew [metric] from X to Y between [dates], measured in [source].”
  3. Add the context a buyer needs: the starting point, what you changed, how long it took.
  4. Put the numbers in HTML text, with the screenshot next to them, not instead of them.
  5. Give each case study its own URL with a descriptive slug and a visible “updated” date.
  6. Link to it from the service page it proves.

Quotable proof is one layer of our AI search visibility work.

04 / OUR OWN SITE

What we saw on prolinkage.com.

05 / MEASURE

How to measure it.

  • Case studies passing all four checks: count them each quarter.
  • Cited proof: ask the engines about your company by name each month and note whether a case study URL appears in the sources.
  • Search demand: Search Console impressions on case-study URLs, monthly.
07 / SOURCES

Sources.

  • Aggarwal et al., the GEO study (KDD 2024), arXiv 2311.09735, v3, 28 June 2024.
  • Prolinkage AI self-test (ChatGPT, Perplexity, Google AI Overview, one run each), 6 November 2026.
  • FullTilt Team Development, Search Console Generative AI features impressions, last 3 months vs previous 3 months (registry FT-04), captured 4 September 2026.
AI Shortlist Plays

The self-check shows what you can see yourself.

What the engines say across repeated runs, which sources they read and who they name instead is harder to see from inside. Our free AI Visibility Snapshot runs a sample of your buyers’ questions and is reviewed by a person.

More than a decade of experience, across businesses and markets.