How do you track your brand in AI answers?

Build a fixed panel of buyer questions, run them in ChatGPT, Perplexity, Google AI Mode, and Claude on a schedule, and log four things for each answer: were you mentioned, were you cited with a link, who else got named, and was it accurate. A spreadsheet is enough to start. The panel shows you shortlists your analytics will never see.

Why track AI answers at all?

Because buyers ask AI for shortlists before they ever visit your site, and you can't see that happen.

TechCrunch reported on Oct 5 that ChatGPT has 1.2 billion weekly users. Some of them are your buyers asking "what are the best tools for" whatever you do. If the answer names three competitors and leaves you out, you lost a shortlist with no visit, no session, and no lead to count.

That's why I treat this as a Signals job. It's the market telling you what it thinks you are, and Signals is where you measure and fix that before you spend more on Convert.

Which prompts go in the panel?

Twenty to thirty prompts, written the way buyers actually type, split across four types.

Prompt type	Example
Category	"best website visitor identification tools for B2B"
Problem	"how do I find out which companies visit my website"
Comparison	"RB2B vs Vector"
Brand	"is [your product] good for [use case]"

Pull the wording from places buyers already use it: sales call notes, support tickets, G2 category names, and the free-text answers from your self-reported attribution field.

Then freeze the panel. If you change the prompts every week, the trend means nothing.

What should I log for each answer?

One row per prompt, per engine, per run:

• Date and engine
• Mentioned (yes or no) and position in the list
• Cited with a link (yes or no, plus the URL)
• Competitors named
• Accuracy (note anything wrong about your product)
• Source pages the engine cited

The last column does the most work. It tells you which pages the engines trust for your category: your blog, a G2 page, a Reddit thread, a competitor's comparison page. That list is where you go to work.

How often should I run it?

Weekly for category and comparison prompts, monthly for the rest.

Answers change from run to run, so run each high-value prompt more than once and note when the answer moves. Use a fresh, logged-out session where you can, so your own history doesn't shape the result.

Why am I missing from AI answers?

Check four causes, in this order.

1. Crawlers are blocked. OpenAI's crawler docs say sites that opt out of OAI-SearchBot won't be shown in ChatGPT search answers, though they can still show up as navigational links. GPTBot is a separate setting. Check your robots.txt for both, plus PerplexityBot and Googlebot.
2. No page answers the question. If buyers ask "RB2B vs Vector" and you don't have a page that answers it plainly, the engine cites someone who does. My own RB2B vs Vector post is built to answer exactly that question.
3. Third parties don't mention you. Engines lean on review sites, comparison posts, and forums. G2 category pages show up again and again.
4. Your pages are vague. If your page can't say what you do in one sentence, an engine can't quote it.

For Google specifically, Search Central says SEO best practices still apply to AI Overviews and AI Mode, with no additional requirements or special optimizations. Good pages that answer real questions are the whole play. The AEO guide for B2B blogs covers how to write them.

What do I do with the results?

Fix what the log points at, then check it against pipeline.

• Wrong facts about you: fix the page the engine cited, and fix your own page.
• A competitor's comparison page gets cited: write the honest version. A comparison page that converts shows the tradeoffs and tells buyers when to pick the other tool.
• Mentioned but never linked: put a plain, quotable answer at the top of the page that should get the link.
• Then join it to meetings. When "ChatGPT" or "Perplexity" starts showing up in how-did-you-hear answers from booked meetings, the panel and the pipeline agree.

A citation is a signal. A meeting is the result. Don't confuse the two.

Should I buy an AI visibility tool?

Maybe later. Run the spreadsheet for a month first so you know which prompts matter and what you'd want a tool to automate. Buying software before you know the questions gets you a dashboard of the wrong prompts.

What should I do this week?

1. Write 20 prompts across the four types.
2. Run them in ChatGPT, Perplexity, Google AI Mode, and Claude. Log every answer.
3. Check robots.txt for OAI-SearchBot, PerplexityBot, and Googlebot.
4. Pick the three prompts where a competitor is named and you aren't. Fix or write the page that should answer each one.
5. Tag AI assistants in your how-did-you-hear answers so you can match them to meetings.

If an AI names three competitors and leaves you out, you lost a shortlist you'll never see in analytics. A spreadsheet and 20 prompts will show it to you.

Start Signals, Convert, Grow

FAQ

What is AI answer tracking?

Running a fixed set of buyer prompts in AI assistants on a schedule and logging whether your brand is mentioned, cited, accurate, and named next to which competitors.

Do I need to allow GPTBot to show up in ChatGPT search?

OpenAI's docs say OAI-SearchBot is the crawler for ChatGPT search, and GPTBot is a separate, independent setting. Allow OAI-SearchBot if you want to appear in ChatGPT search answers.

How many prompts should I track?

Start with 20 to 30. Enough to cover your category, main problems, top comparisons, and brand questions. Small enough to run by hand.

Is AI answer tracking the same as SEO?

They overlap. Google says normal SEO best practices apply to its AI features. The tracking is different: you're logging answers and citations, not just rankings.