How it works

From a question to an answer you can act on

The mechanics are simple. The part that takes judgement is reading the result, so this page covers both.

Start freeRead the methodology

14-day trial, no card required

A run, end to end

Four steps, repeated daily.

  1. 1

    Pick the questions your buyers ask

    Not keywords. Whole questions, phrased the way a person types them into an assistant. We suggest a starting set from your category and competitors, and you edit it. Ten to thirty questions is enough to see a shape; the number matters less than keeping them fixed once chosen.

  2. 2

    We run them on a schedule

    Every question goes to every engine, daily by default. Same prompt, same location, so a change in the answer means something changed in the world. Runs are queued and retried, and a failure leaves the denominator rather than counting as an absence.

  3. 3

    We extract what happened

    For each answer we record which brands were named and in what order, which URLs were cited, and the full text. Your competitors are tracked alongside you by default, because a share of voice with nobody to share it with is just a number.

  4. 4

    You read it per engine

    Six columns, never one score. The screen shows how often you were named, how often you were cited, and the sample behind both, so the first thing you learn is which engine to work on.

Reading the result

Three problems that look identical

Not being mentioned is not one problem. It is three, and they need different work. Telling them apart requires the citation data, which is why we store sources rather than only counting names. It shows up most clearly on Perplexity, which cites more heavily than any other engine, and on Google AI Overviews, where the citation panel and the prose are visibly separate.

Never retrieved.
Your pages are not in the citation list on any answer. The assistant is not reading you at all. This is an indexing and content-coverage problem, and more brand mentions elsewhere will not fix it.
Retrieved, not cited.
You appear occasionally but rarely enough to look like noise. Usually a depth problem: the page touches the topic without answering the question the way the model needs it answered.
Cited, not named.
You are in the sources on many answers and named in almost none. The model reads you and recommends someone else. This is a framing problem, and it is the one people most often misdiagnose as needing more content.
States · ChatGPT · 30 answers
Named
You were recommended by name.
14 / 30
Cited, not named
Your page was read. Someone else got the credit.
7 / 30
Retrieved, not cited
Fetched, then not used.
4 / 30
Not retrieved
The assistant never saw you.
5 / 30
Illustrative figures from a demo workspace, not a customer's data.

Questions about running it

How long before I see anything useful?
The first answers land within minutes of your first run, and you will see a shape the same day. A trend takes about two weeks, because answers vary between runs and it takes that long for a direction to separate from noise.
How many questions should I track?
Ten to thirty to start. Enough to cover the ways your category gets asked about, few enough that you keep them stable. Consistency matters more than breadth: a question you rewrite each month produces no trend at all.
Do I need to add my competitors?
You do not have to, but the product is much less useful without them. Being named in two answers out of twenty means one thing if nobody else is named either, and something very different if one competitor is named in eighteen.
What happens when a run fails?
It is retried, and if it still fails it is excluded from your denominator rather than recorded as an answer that did not name you. You are emailed if a whole run fails or if most of one does, because a silent failure is indistinguishable from a quiet week.

The rules behind every number are in the methodology, and the engines are listed on platforms.

See where you stand

Connect a brand, pick the questions your buyers ask, and get your first snapshot in a few minutes. No card for the trial.

Start freeRead the methodology