ChatGPT · answer coverage

What ChatGPT says about your store, on the record.

Asked on a fixed rotation, kept answer by answer, and read over a 21-day window.

Engine ChatGPT

900M+ people use it every week. OpenAI · Feb 2026

The most used AI assistant, and the first place a buyer asks about a product, a brand, or a store.

Representative demo · Threshold Footwear
Representative demo · Threshold Footwear
Representative demo · Threshold Footwear

Why it matters

Why ChatGPT is worth measuring.

01

The largest audience in AI

No assistant is asked more questions by more people. When a buyer asks an AI about your category, this is usually the one.

02

Shopping is built in

ChatGPT answers product questions with named stores and links, and checkout is moving inside the chat. Discovery here is an answer, not a results page.

03

An answer has fewer places than a list

A results page has room for ten stores. An answer names two or three. You are in it or you are not.

What Burnish does

Four things, on this engine, every cycle.

Nothing here is a projection. Each row is a reading, a change, or a comparison between the two.

Ask

The questions your buyers actually ask

A board of buyer questions in sentences, not keywords — 50 on Presence and Authority, 100 on Omnipresence.

Representative demo · Threshold Footwear
Representative demo · Threshold Footwear

Check

What it says when asked about you

A 12-question battery on legitimacy, returns, shipping, warranty and provenance, and whether ChatGPT can answer each one.

Representative demo · Threshold Footwear
Representative demo · Threshold Footwear

Fix

What the answer is drawn from

Descriptions, structured data, FAQs and alt text written back to your catalog — every change approved first, and reversible in one click.

Representative demo · Threshold Footwear
Representative demo · Threshold Footwear

Prove

What moved after the change

Readings before and after, on the same questions. Attribution is correlation over a window, and the page says so.

Representative demo · Threshold Footwear
Representative demo · Threshold Footwear

Under the hood

How Burnish reads ChatGPT.

Sampler · ChatGPT GPT-5.4 mini

“Is Threshold Footwear legit?”

I couldn’t verify Threshold Footwear as an established retailer.

Doesn't know you

Read over a rolling 21-day window

Model queried

GPT-5.4 mini. When a provider changes the model behind an assistant, the page changes with it rather than keeping a name that is no longer true.

What gets asked

A 50-question board of buyer questions on Presence and Authority, 100 on Omnipresence, plus a 12-question reputation battery covering legitimacy, returns, shipping, warranty and provenance.

How often

The reputation battery runs on a 72-hour cycle. Discovery questions fire on a rotating schedule — one every 120 minutes on Presence, 90 on Authority, 35 on Omnipresence — so the board is swept continuously rather than in one daily batch.

How it is scored

Answers from ChatGPT are read over a rolling 21-day window, and a score is withheld below ten samples rather than shown as a number that a handful of readings cannot support.

Questions.

Does Burnish change what ChatGPT says about my store?

Not directly. It changes what your store publishes — descriptions, structured data, FAQs — and then measures whether the answers change. No one can write into a model.

How quickly does a fix show up in answers?

Readings continue on the same rotation after a change, and the 21-day window is what a score is read over. Burnish reports when the reading moves; it does not promise a date.

Is ChatGPT included on the entry plan?

Yes. Presence samples ChatGPT, Perplexity and Gemini. Claude and Grok start on Authority.

Start here

Ask it about your own store first.

The free audit reads your catalog and shows what an assistant has to work with.