How to Track Your Brand Across AI Engines

Your buyers are asking ChatGPT and Perplexity for recommendations before they ever reach your site. Each engine names a few brands and skips the rest. This guide shows you how to find out whether you're named, across the engines that matter, and what to do about the gaps.

Why this matters

A search engine shows you your rank if you check. An AI engine doesn't. It gives a different answer to every person who asks, keeps no public scoreboard, and never tells you when it recommended a competitor instead of you.

So the brands losing visibility inside AI answers usually don't know it's happening. The first step isn't optimization — it's measurement. You need a baseline before you can tell whether anything you change is working.

What to track

Three signals do most of the work:

• Mentions — how often an engine names your brand, with or without a link. This is your share of attention.
• Citations — how often an engine links to your site as a source. A mention with no citation means the engine knows you but sent the click somewhere else.
• Share of voice — how often you're named versus your competitors for the same questions. "Named 4% of the time" only means something next to a rival at 20%.

Track each per engine and per country. ChatGPT might name you often while Gemini ignores you, and you might lead in the US but trail in the UK. An average across all of them hides the gap you most need to see.

A simple step-by-step

1. Pick prompts your buyers actually ask. Skip your brand name. Buyers researching options don't search for you — they ask open questions like "best [category] tool for small teams" or "[Tool A] vs [Tool B]". Write 10 to 20 of these. Comparison and "best X" prompts matter most, because those are the ones that produce a named shortlist.

2. Run each prompt across every engine. Ask the same questions in ChatGPT, Claude, Perplexity, Gemini, Google AI Overviews, and Copilot. Record who gets named, who gets cited, and which URL each citation points to. Doing this by hand is slow and the answers shift between runs, which is the main reason teams automate it.

3. Read the gaps. Look for three patterns. Engines where you're never named. Engines where you're named but a competitor is cited instead of you. And countries where your share of voice drops. Each pattern points to a different fix.

4. Act on what you find. If you're absent from the "best X" answers, you likely need content that directly answers those questions in a citable, well-structured way. If you're mentioned but not cited, the engine knows you but trusts another source more — so strengthen the page it should be linking to. Then re-run the prompts and compare against your baseline.

How Geotally does this

Geotally runs steps two and three for you. You add the prompts your buyers ask, and it tracks whether six engines — ChatGPT, Claude, Perplexity, Gemini, Google AI Overviews, and Copilot — name and cite your brand, broken down by country, with share of voice against the competitors you choose.

A weekly digest shows what moved, so you're not re-running prompts by hand or guessing whether last month's change helped. The free tier covers a starter set of prompts, which is enough to see where you stand before deciding what to fix.

If you want the background on the category first, see our explainer on what Generative Engine Optimization is. When you're ready for more prompts, countries, or competitors, the pricing page lays out every plan.

Start with a baseline

You can't improve what you haven't measured, and AI engines won't hand you the numbers. Pick your prompts, run them across the engines your buyers use, and see who gets named. Once you have a baseline, every change you make has something to be measured against.

Start free and see how AI engines describe your brand today.