Solution

From one published URL to an auditable scorecard.

The pipeline, the rules the numbers obey, and every metric on the scorecard — in the detail you’d want before trusting it.

01 · How it works

From one published URL to an auditable scorecard.

/01

Paste a URL

No tracking code, no CMS plugin. Your published page — article, product, guide, docs — is the only input.

/02

Real questions

We prepare the questions your target audience will ask about the topic — from broad to your exact angle — and you can add your own.

/03

Probed, with repeats

Every engine answers every query several times. One answer is an anecdote; repetition is a measurement.

/04

A scorecard you can audit

Raw answers and page snapshots are kept on disk. Every number traces back to the evidence — and recomputes from it.

02 · Principles

Two rules, baked into every number.

Rule one

Two levels, never one.

The rate for this URL and the rate for any of your properties are always shown together. You often win a query with a different page — tracking a single URL undercounts your real position.

This URL
34%
Any property
61%

The 27-point gap is your sister titles doing the winning.

Rule two

Honest about influence.

When your article is cited, we compare the AI’s claims against what you actually wrote and score fidelity with confidence. When it isn’t cited, we never claim it “influenced” the answer — experimental heuristics stay out of the headline numbers.

citedRepresentation scored against your text9.1/10
not citedInfluencenot claimed

A tool you can put in front of your board without an asterisk.

03 · The scorecard

Every angle of the answer, on one page.

a.

Who got cited instead

On the queries you lost: the competitive share, brand by brand, domain by domain.

b.

Prominence, not just presence

Woven into the prose or parked in a link list? First position or eighth? It changes what a citation is worth.

c.

The specificity gradient

Where you start winning — broad category questions, or only when the query names your exact angle.

d.

Engine vs engine

The same prompt set across every assistant, side by side — where you’re strong and where you’re invisible.

e.

Representation fidelity

When you are the source: does the answer say what you actually wrote — or a distorted version of it?

f.

Citation trend

Recurring runs turn the scorecard into a series: rates, deltas, and the moment a rival takes your spot.

See how your story is being ToldBy AI.

Start with a published URL, or talk to us about measuring a brand campaign across a defined set of questions and engines.

Get your citation reportBook a demo