Why does a sitemap matter for SEO and GEO?

Fihrist methodology

A measurement is useful only when its limits are visible.

Fihrist compares like with like, keeps the evidence attached, and never presents one changing answer as a universal truth.

Monochrome geometric artwork representing Fihrist AI visibility research

Fihrist begins with a defined brand, market, competitor set, topic scope, and buyer intent. Answers are retained at engine and question level. Rates keep their sample context, and repeated measurement is used to separate a meaningful pattern from ordinary output variation.

AI answers are variable. The method must account for that.

A screenshot can be useful evidence, but it is not a market-wide conclusion. Reliable interpretation depends on comparable prompts, declared scope, retained answers, sample context, and honest confidence language.

What you can learn

Measurement conditions

Prompts are written in Turkish, measurement runs from a Turkey location, and results are checked against the real apps on a weekly cadence. Every report states the conditions it was measured under.

Defined scope

Document the brand, competitors, engines, topics, intents, period, and market before measuring.

Comparable runs

Use the same evaluation frame when comparing brands and engines.

Answer-level evidence

Keep the response, mention, recommendation, and source behind every finding.

Qualified confidence

Show sample context and limitations instead of false precision.

How the analysis works

Define the market

Agree on the brand, competitors, audience, and measurement boundaries before measurement begins.

Collect comparable evidence

Run the same buyer questions across the same AI engines, with every finding linked to the answer behind it.

Turn the gap into work

Turn the findings that matter into source, content, brand, or technical priorities, then measure again.

fihrist.io/dashboard/insights
Search…⌘KExport

Visibility

İİyi Sigorta
All modelsRun report

Visibility, share of voice, and citing sources are calculated from real engine runs.

Visibility

İyi Sigorta across ChatGPT, Gemini, Google AI Overviews, Perplexity, and Grok

38%±5% range+16 pts

İyi Sigorta appeared in 38% of the 320 sampled AI answers.

Visibility ranking

By mention frequency

BrandVisibilitySOVPerception
1İyi SigortaYOU38%34.481
2Pera Sigorta31%27.974
3Emir Sigorta22%18.669
4Zenka Sigorta14%11.277
5Mira Sigorta8%5.372
6Armut Sigorta3%2.666

Share of voice

Brand and competitors

İyi Sigorta34.4%
Pera Sigorta27.9%
Emir Sigorta18.6%
Zenka Sigorta11.2%
Mira Sigorta5.3%
Armut Sigorta2.6%

Average position

Position in answers that mention you

#1.60.3

Live product demo with the fictional İyi Sigorta account. Hover the charts; the demo also cycles through views on its own.

What your team can do with it

  • Know exactly what a result represents
  • Audit every important finding
  • Avoid conclusions from isolated answers
  • Compare change on a consistent basis

Frequently asked questions

Why can the same prompt return different answers?

Models are probabilistic and their systems, sources, and context change. Variation is expected and must be reflected in interpretation.

Does a visibility score predict revenue?

No. It is an observable indicator within a defined answer set, not a direct revenue attribution model.

How do you communicate confidence?

By retaining evidence, stating scope and sample context, repeating comparable measurements, and qualifying findings that are not stable.

Related pages

See where your brand stands in AI answers.

Request early access. We will help you define the market, competitors, and buyer questions worth measuring.

Request early access