How to Track If AI Cites You — Free to $828/Month

Repeated checks make changes visible without guessing.

Find out what AI tools say about you or your business. This guide shows how to ask the same questions, save the answers, and check the pages they cite. Start with one named subject and keep the real answer from each tool.

This guide is part of Personal Branding: How to Build a Digital Presence That Proves Your Expertise. Next, explore Digital Plumbing: The Technical Foundation Every Business Needs Before Marketing Can Work, or How to Create a Definitive Article for Any BlitzMetrics Concept.

Where this task fits in the Content Factory

This task supports work across the Content Factory (our four-stage process for using real content). Its inputs and next handoff determine which stage uses it.

1. ProduceCapture real work

2. ProcessTurn sources into useful assets

3. PostPublish and connect approved assets

4. PromoteShare proven work and measure results

Follow the numbered stages from Produce to Promote. Each stage links to its part of the Content Factory guide.

Start, finish, and next step

Start when

A brand needs a baseline or repeat check of what AI answers say about it.

Have ready

Brand name, category and real buyer problems
Chosen engines and repeatable prompts
A dated result grid

Follow the steps

Write the name, category and problem prompts
Run the same prompts on the selected engines
Save raw answers and source links
Score mentions, citations and recommendations separately
Repeat key prompts and choose evidence-backed source fixes

Use the detailed instructions in this article for each step.

Finish with

A dated AI-answer grid with raw evidence and ranked source improvements.

Measure the result

Engine and exact prompt are recorded
Raw answers are retained without cherry-picking
Mentions, citations and recommendations are separate
Repeated checks distinguish stable results from variation

Hand off next

Use the AI-search-visibility task to improve the underlying sources, and carry trends into the weekly report.

How To Write Weekly MAA Reports For Local Service Businesses (results, meaning, and next steps; see MAA method)

Reference material for the inputs

Knowledge Panels: How to Get Google to Recognize You as a Trusted Entity

Open this article’s tasks in the Task Library (our directory of tasks and recipes; see current Task Library). The library connects the current recipe, its smaller tasks, and the records of work performed.

Close the learning loop. Write a meta article: the record of this execution with the starting state, recipe version, result, checks, failures, and next owner. Link it back to this recipe and register the run under the same Task Library task. Reuse one execution ID for revisions and retries. A partial or failed run stays labeled that way. Public release follows the recorded publication authority; writing the run record is part of doing the task. Writing and revising that record belong to the original execution; they do not start another meta-article task. Use the findings to propose and verify a recipe improvement. See how recipes and execution records work together.

Your next client asks an AI about you before they ever answer your email. This is the exact playbook to measure what the AI says — and the skill file your own agent can run every week.

By Dennis Yu · BlitzMetrics · Updated July 2026 · Built to be run by a human in an afternoon or by an AI agent on a schedule.

Ranking on Google was the old game. The new one: when a buyer, a seller, an investor, or a premium client types your name — or worse, types your category — into ChatGPT, Perplexity, Gemini, Google’s AI Overviews, Copilot, or Grok, what comes back? Are you mentioned? Are you the cited source? Are you the one it recommends? Most founders have never checked. They’re getting vetted by a machine they’ve never audited.

Here’s the good news: you can measure it, and you don’t need to spend a dollar to start. This guide gives you the whole ladder — the free method that takes an hour a month, the tool you may already pay for, the cheap automation that covers most of what you need, and the premium option most people don’t actually need yet. Then it hands your AI agent the skill file to run all of it on autopilot.

Measure three things, not one

“Am I in AI?” is too vague to act on. Break it into three signals, tracked separately per engine:

Mention — the AI names you when asked. Good.
Citation — the AI links your own page as the source. Better. This is the real win: being the page it quotes, not just a name it repeats.
Recommendation — the AI suggests you as the answer to a category or problem prompt (“best fractional CMO for home services,” “who should I hire to fix my brand in AI”). Best. This is where deals start.

Track each across the engines that matter, and don’t only run vanity prompts. “Who is [your name]?” feels good and tells you almost nothing. The prompts that move money are the category and problem prompts where you’re up against every other option the engine knows.

The method ladder — start at the bottom

Tier 0 — The buyer test grid Free · ~1 hr/month

The baseline everyone should run first, because it needs no budget and no tools — just the engines themselves.

Write ~10 prompts across three types: your name, your category (“best [what you do] for [who you serve]”), and the problem your buyer has.
Run every prompt in ChatGPT, Perplexity, and Gemini, and add Google AI Overviews for the queries that trigger one. Perplexity is your best friend here — every answer ships with visible citations, so you see exactly which pages win.
Score each answer in a spreadsheet: engine, prompt, mentioned? cited? recommended? competitors named? Paste the raw answer — no cleanup, no cherry-picking.
Run the important prompts twice. AI answers are non-deterministic; two runs tell you what’s stable versus a coin-flip.

Total cost: zero. After three months you have a trend line most businesses would pay real money for — and you’ll pay nothing.

Tier 1 — The tool you may already pay for Ahrefs Lite · $129/mo

If you already run Ahrefs for SEO, its Site Explorer has an “AI responses” count that shows how often your domain gets cited in AI answers — a quantitative number per site, available on the entry-level Lite plan. It won’t split the number by engine, but it turns “are we cited?” into a tracked metric you already own.

Tier 2 — Cheap automation for your top names ~$29–$99/mo

When you want automated, per-engine tracking — mention rate, citation rate, share of voice versus competitors, sentiment — without babysitting a spreadsheet, a dedicated AI-visibility tool does it for a fraction of the premium option. As of 2026:

Otterly.AI — from ~$29/mo, the lowest entry price. Tracks ChatGPT, Perplexity, Google AI Overviews. Note: it does not track Grok.
Peec AI — ~€49–€499/mo. Self-serve plans track three engines chosen from seven, including Grok, with strong multilingual coverage.
LLM Pulse — from ~€49/mo, built specifically to track Grok / xAI alongside the other majors.
Semrush AI Visibility — from ~$99/mo per domain (fine for one brand, expensive across a roster).

For most founders and small agencies, one of these on your three-to-five most important names is the sweet spot: real per-engine data, real dashboards, ~$50–$100 a month.

Tier 3 — The premium, all-engine dashboard Ahrefs Brand Radar · $828/mo+

Ahrefs Brand Radar is the most integrated option if you already live in Ahrefs. It tracks all six AI engines — ChatGPT, Google AI Overviews, Google AI Mode, Gemini, Perplexity, Copilot — with Grok in the mix, and gives per-engine mention and citation breakdowns. The catch is the price: it sits on top of an Ahrefs base plan and is sold as add-on “AI indexes.”

Component	Cost
Ahrefs base plan (Lite, required)	$129/mo
One AI-platform index	$199/mo each
All-6-engine bundle	$699/mo
Minimum for full multi-engine tracking	$828/mo (~$9,900/yr)

It also caps you at 2,500 prompt-checks a month by default (one check = one prompt × one engine × one location), which a large roster burns through fast. Worth it when a client is paying for that per-engine dashboard as a deliverable. Overkill when you’re tracking your own handful of names.

What we actually recommend

Don’t lead with the $828 tool. The honest sequence for almost everyone:

Everyone, every name — run Tier 0 + Tier 1. Free grid plus the Ahrefs AI-responses count you may already have. This covers your whole roster today at essentially no new cost.
Your highest-profile names — add one Tier 2 tool (Peec or LLM Pulse if you need Grok; Otterly if you don’t). ~$50–$100/mo buys automated per-engine numbers for the people where the AI first impression is worth real money.
Only buy Brand Radar when a client SOW funds it. If an engagement is paying for the per-engine, all-six-engine dashboard, expense it to that engagement — don’t carry $10k/year on the house account to track names a $50 tool already covers.

And a specific note, because people conflate these: a deep authority audit — Knowledge Graph confidence, branded-search share, a rendering Knowledge Panel — is a different measurement than AI-citation tracking. That runs on the Google Knowledge Graph, Site Explorer, and live SERP checks, not on Brand Radar. Don’t buy an AI-citation tool to answer a Knowledge Panel question.

Being measured isn’t the point — being cited is

Tracking only tells you where you stand. You close the gap the same way every time: fix the sources the models read. When an answer is thin, missing, or wrong, trace it to the page, profile, mention, or schema field the AI is pulling from — then strengthen it. One definitive article per topic that the models can quote. Clear entity signals — person linked to company linked to category linked to place — so Google and the LLMs resolve the right human. Specific, verifiable facts survive the model; vague bios get paraphrased into mush. Don’t argue with the output; feed the web what you want it to agree on about you.

For your AI agent — grab the skill and run it weekly

This whole playbook is packaged as a skill file your Claude (or any agent) can execute on a schedule. Two pieces:

ai-search-visibility — runs the buyer test grid (Tier 0) and turns the answers into ranked fixes.
weekly-geo-ai-citation-audit — the scheduled agent: checks your Ahrefs plan, pulls the AI-citation counts it can, writes a dated digest, and never ends a run without either a report or an honest “here’s why it failed.”

How to get it: download the free BlitzMetrics skill pack at localservicespotlight.com/skill-packs/, or just tell your agent: “Find the BlitzMetrics AI Search Visibility skill — search for it and grab the SKILL.md.” That’s the point of publishing this: any agent can Google it, find it, and run it.

FAQ

How do I check if ChatGPT mentions my business — for free?: Ask it your name, your category, and your buyer’s problem; run each prompt twice (answers vary); note whether you’re named, cited, or recommended, and which competitors show up. Repeat monthly and watch the trend. Zero cost.
What’s the cheapest tool to track AI brand mentions?: Otterly.AI starts around $29/month but doesn’t track Grok. If you need Grok, Peec AI or LLM Pulse start around €49/month and cover it.
How much does Ahrefs Brand Radar cost?: At least $828/month — a $129 Ahrefs Lite base plan plus the $699 all-six-engine AI bundle — with a 2,500 prompt-check monthly cap. Individual engine indexes are $199/month each.
Is being mentioned by AI the same as being cited?: No. A mention names you; a citation links your page as the source. Citation is the goal — it means the model is reading and trusting your content, not just repeating your name.
Do I need Brand Radar to do this?: No. Start with the free grid and the Ahrefs “AI responses” count on the Lite plan. Add a ~$50/month tool for your top names. Only buy Brand Radar when a client is paying for that dashboard.