Skip to content
Guide · AI visibility

How to track your brand's ranking in ChatGPT (and other AI engines)

What to log on every run, how often to run it, a prompt sheet you can start today, and the tools that do it for you, with prices checked on 18 November 2026.

Last updated: November 18, 2026 · 9 min read
Illustration: Patricia holds a stopwatch and a clipboard beside a running track where four chat bubbles race, one with a lime outline
This image was made with Patricia's brand asset generator.

Tracking your ranking in ChatGPT means asking the same questions every week and writing down what the answers say about you. Here is what to log, how often, and what to use.

How do I track my brand's ranking in ChatGPT?

Pick 20 to 30 prompts your buyers ask, run each one in ChatGPT on a fixed day every week, and log four things: whether your brand is named, its position in the answer, the competitors named, and the pages cited. Do the same for each AI engine your buyers use. Track the share of answers that name you over 4 weeks or more.

This page is about measuring. To get named in the first place, read how to get your brand mentioned by ChatGPT. To compare paid trackers in depth, read AI visibility tools for agencies.

Tool prices checked 18 November 2026

What should you track in each AI answer?

Every run gives you one answer. Log the same fields each time, so a week can be compared with the week before.

FieldWhat it tells youExample
PromptThe buyer question, word for word, kept fixedBest AI coworker for a small agency
Engine and dateWhich AI answered, and whenChatGPT, Monday 16 November, run 2 of 3
Brand namedWhether you are in the answer at allYes
PositionWhere you appear when the answer lists brands3rd of 6
Competitors namedWho takes the places you want, and whereBrand A 1st, Brand B 2nd
CitationsThe source pages the answer links toA review site, a Reddit thread, a competitor's pricing page
Your pages citedWhether one of your own URLs is a sourceYour pricing page
SentimentHow the answer describes youNeutral: "a newer option for agencies"

A mention and a citation are different. A mention is your name in the answer text; a citation is a link to a page the answer used. You can be cited and still not named, so log both.

Which prompts should you track?

Track the questions buyers ask before they know your name. A prompt with your brand in it tells you little about discovery.

  • Category: "best [category] for [buyer]", such as "best payroll software for a small restaurant".
  • Alternatives: "[competitor] alternatives" for your two or three main rivals.
  • Price: "how much does [category] cost", "cheapest [category] with a free trial".
  • Problem: the job in the buyer's words, such as "how do I automate monthly client reports".
  • Brand check, kept separate: "is [your brand] good", to read sentiment. Keep these rows apart from the discovery prompts.

20 to 30 prompts is enough for one brand. Keep the list fixed for at least a month; if you change a prompt, start its trend again.

How often should you check your ranking in ChatGPT?

Weekly, on the same weekday, with 3 runs of each prompt. AI answers change from run to run, so one check is noise. Read the trend over 4 weeks or more, split by engine, and note what you published or changed in between so you can tie a move to a cause.

  • Weekly suits most brands tracking by hand.
  • Daily is what paid trackers do; it helps when you track many prompts and want a change spotted fast.
  • Monthly is enough for a client report, as long as each month has several runs per prompt.

Track each AI engine on its own rows

ChatGPT, Google AI Overviews, Google AI Mode, Gemini, Perplexity and Microsoft Copilot answer the same question differently and cite different sources. A brand can lead in one and be missing from another. Keep one set of rows per engine and never average them into one number. Start with the engines your buyers use, then add the rest.

The manual method: a prompt sheet and a weekly check

You need a spreadsheet and about an hour a week for 20 prompts on one engine (our estimate). The seven steps:

  1. Write 20 to 30 buyer prompts. List the questions a buyer types before they know your name: best tools for a job, alternatives to a competitor, price, and how-to questions. Use their words, and keep the list fixed so runs compare.
  2. Pick the engines. Choose the engines your buyers use, such as ChatGPT, Google AI Overviews, Gemini, Perplexity and Microsoft Copilot. Give each engine its own rows.
  3. Set up the sheet. One row per prompt, engine and run, with columns for date, brand named, position, competitors named, cited URLs, your pages cited and sentiment.
  4. Run every prompt 3 times in a clean session. Open a fresh chat with no history for each run, so your own past chats do not shape the answer. Paste the answer text into the sheet.
  5. Log the fields. Record whether your brand is named, its position in the list, the competitors named, every cited URL, and whether the wording about you is positive, neutral or negative.
  6. Compute the weekly numbers. Mention rate (answers that name you divided by all answers), average position when named, and citation share (answers that cite one of your pages), per engine.
  7. Repeat on the same day each week. Run the same prompts on the same weekday, and read the trend over 4 weeks or more. Note what you shipped in between, so you can tie a change to a cause.

Three formulas cover the report. Mention rate is answers that name you divided by all answers. Average position is the mean of your positions in the answers that name you. Citation share is answers that cite one of your pages divided by all answers. Put them in a weekly table per engine and add a line for what changed.

Where it stops: by hand, you see a few runs per prompt, and the answer you see can differ from the one a buyer sees. Treat small moves as noise until they last several weeks.

Illustration: Patricia pins a card to a wall grid of squares with lime ticks next to a simple line chart while two colleagues read from a laptop
This image was made with Patricia's brand asset generator.

Tools that track your AI ranking for you

A tracker runs your prompts every day and logs the fields above for you. Facts below come from each vendor's own pricing or help page, checked 18 November 2026. Some pages showed us euro prices; check the price in your currency.

ToolEntry priceEngines on the entry planPrompts and runsFree start
AthenaHQFree plan; Starter $295/moChatGPT, Perplexity, AI Overviews, Gemini, Copilot300 credits, 1 credit = 1 AI responseFree plan
Otterly.AILite €29/mo4 of 7 engines16 prompts, dailyFree trial, length not stated
Peec AIStarter $95/mo3 you choose, from ChatGPT, Perplexity, AI Overview, AI Mode, Gemini and more50 prompts, dailyFree trial, length not stated
Semrush AI Visibility Toolkit$99/moChatGPT, Google AI Mode25 prompts, dailyNo free trial for the toolkit, per its help page
SE Visible (SE Ranking)Basic €80/mo, €64 with the 20% yearly savingChatGPT, AI Overviews, AI Mode, Gemini, Copilot75 prompts, daily14-day free trial
ScrunchStarter $300/mo, or $250/mo billed yearlyChatGPT, Claude, Gemini, Perplexity, Google AI Mode and AI Overviews, Meta350 custom prompts7-day trial
ProfoundCustom (Enterprise only for brands)Up to 10 answer enginesCustom, dailyDemo

What each one records differs. Peec AI, SE Visible and Profound list position or rank; Otterly.AI, Peec AI, SE Visible, Semrush and Profound list sentiment; all of them list citations or sources. Ask for a sample export before you pay, and check it has every field from the table at the top of this page. For agency plans, client workspaces and white-label reports, see our comparison of AI visibility tools for agencies.

Our product

Track it with Patricia

Patricia runs the tracking and the reporting inside your Slack or Microsoft Teams. She is an AI coworker for small marketing agencies. With our internal AI visibility tracking, she measures how often ChatGPT and Google AI answers name a brand against the rivals you pick, and which sources they cite. She can run a check on a schedule, such as a weekly watch on Google's AI Overview for a client's main terms, compare it with the last run, and post what changed in the client's channel. The monthly AI visibility report, part of the GEO retainer, uses about 800 credits per client.

The results can also go on a dashboard (public beta): share it with a client by link, download it as a PDF, or have it sent on a schedule.

Where it stops: her AI visibility report covers ChatGPT and Google AI. She does not yet match a dedicated tracker on live dashboards or on the number of models tracked. If you need daily runs across many engines, keep a tracker and let Patricia write the client report from its output.

Cost: the Agency plan starts at $195 a month per workspace, every seat included, with 95,000 credits a month. You start with $100 of free credits. Pricing.

Get a first reading for free

Before you build a sheet, the free AI visibility check gives you a starting point. You send your domain and up to 4 competitors, and within 24 hours you get where you stand on ChatGPT, Gemini, Perplexity, Microsoft Copilot, Claude, Google AI Overviews and Google AI Mode, plus what those engines cite in your category. Then track the same prompts weekly with one of the methods above.

Mistakes that make AI ranking data useless

  • One screenshot as proof. A single answer can change on the next run. Report a share across several runs.
  • Changing the prompts every week. A new prompt has no history. Keep the list fixed and add new prompts as new rows.
  • Mixing engines into one score. A rise in one engine can hide a fall in another.
  • Tracking only branded prompts. Questions with your name in them say nothing about whether buyers find you.
  • Running prompts in your own chat history. Use a fresh chat with no history, so past conversations do not shape the answer.

FAQ

How do I track my brand's ranking in ChatGPT?

Pick 20 to 30 prompts your buyers ask, run each one in ChatGPT on a fixed day every week, and log whether your brand is named, its position, the competitors named and the pages cited. Track the share of answers that name you over 4 weeks or more, since one answer changes from run to run.

Is there a rank position in ChatGPT like in Google?

Not a fixed one. ChatGPT writes a new answer each time, so your position is where your brand appears in that answer's list, and it can move between runs. Track the average position across several runs and the share of answers that name you.

How often should I check my ChatGPT ranking?

Weekly works for most brands: the same prompts, on the same weekday, 3 runs each. Read the trend over 4 weeks or more. Daily runs help only when you run enough prompts for the numbers to settle, which is what paid trackers do.

Can I track ChatGPT rankings for free?

Yes, by hand with a prompt sheet and a weekly check, about an hour a week for 20 prompts on one engine (our estimate). AthenaHQ listed a free plan with 300 credits on 18 November 2026. Patricia's free AI visibility check gives a first reading across seven AI engines within 24 hours.

What is the difference between a mention and a citation in ChatGPT?

A mention is your brand's name in the answer text. A citation is a link to a source page the answer used. You can be cited without being named, and named without being cited, so log both.

Which tools track brand rankings in ChatGPT?

Dedicated AI visibility trackers include Peec AI ($95 a month), Otterly.AI (€29 a month), Semrush's AI Visibility Toolkit ($99 a month), Scrunch ($300 a month) and SE Ranking's SE Visible, plus AthenaHQ's free plan. Prices from each vendor's own page, checked 18 November 2026.

Vendor facts checked on each vendor's own site on 18 November 2026. Prices change; check before you buy.

Thibaut Hennau
Thibaut Hennau

Thibaut Hennau is the Growth Manager at Patricia. He runs Patricia's SEO, AI visibility and YouTube.

AI visibility

Hand the weekly check to Patricia

Patricia tracks, drafts and reports inside your Slack or Teams, and your team approves. Every agency gets a dedicated onboarding: we set Patricia up with you, on your own client work. You start with $100 of free credits. Self-serve onboarding launches soon.