How to Track If ChatGPT, Claude, and Perplexity Are Citing Your Business (2026 Guide)

Most businesses have no idea whether ChatGPT, Claude, or Perplexity name them — and that is a problem, because you cannot improve a number you have never looked at. This is the rolled-up-sleeves guide to actually measuring your AI citations: how to test by hand, which tools are worth paying for, and how to run a free spreadsheet that turns "I think we're invisible" into a list of fixes. No theory. Just the operator's process.

Why does citation tracking matter in 2026?

Citation tracking matters because AI answers are now a primary discovery surface, and you are either named in them or you are not. There is no "page two" of an AI answer — it cites a few sources and stops. If you are not one of them, you do not exist for that customer, no matter how well you rank in traditional search.

The scale is real. ChatGPT reached roughly 800 million weekly users (OpenAI, 2025), Perplexity handles well over 100 million queries a month (Perplexity, 2025), and Google's AI Overviews now appear on a large share of informational searches (industry analyses, 2025). Yet almost no local business has ever checked whether these engines mention them. That gap is the opportunity and the risk at once: the businesses that start measuring now can see exactly where they are losing and fix it, while their competitors are still guessing. If you want the strategy behind the measurement, start with the complete AEO 2026 guide; this post is purely about tracking.

There is also a defensive reason to track, not just an offensive one. AI engines occasionally describe businesses inaccurately — wrong services, outdated hours, a competitor's detail attributed to you — and an error inside a confident AI answer does real damage, because customers treat it as fact rather than as a guess. The only way to catch those errors is to look for them. I have watched an engine confidently describe a clinic's services using a competitor's list, and the owner had no idea for months. Tracking is how you find both the citations you are missing and the misinformation you need to correct before it quietly costs you customers.

How do you manually test AI citations?

The fastest, cheapest way to test your AI citations is to ask the engines the questions your customers ask and write down who gets named. You do not need a tool to start — you need twenty minutes and an honest list of queries.

Test three kinds of questions. First, industry and service questions — "how much does [service] cost," "is [treatment] worth it" — to see whether you are cited as an authority. Second, "best in [city]" and "near me" questions, which are where local recommendations are won or lost. Third, brand-name searches — "what is [your business]" and "is [your business] any good" — to check how the models describe you, because they sometimes get it wrong, and an inaccurate description is its own problem to fix.

Run each query across all four surfaces that matter: ChatGPT, Claude, Perplexity, and a Google search that triggers an AI Overview. The same question often returns different sources on each, so testing one engine tells you almost nothing about the others.

Here is the systematic process I use with every client.

1. List 20 real customer questions — the actual things people ask before buying, not keyword phrases.
2. Test each across all four engines, one engine at a time, so you can compare.
3. Record who gets named — you, a competitor, or no clear source.
4. Capture the context — the exact sentence and the page that likely fed it.
5. Search your brand name directly to check accuracy of how you are described.
6. Repeat every two to four weeks so you are tracking a trend, not a snapshot.

On cadence: weekly is overkill for most businesses and monthly is the floor. Biweekly is the sweet spot, with an extra pass whenever you ship a significant content or schema change so you can attribute the movement.

Which automated AEO tracking tools are worth it?

A new category of tools now tracks brand mentions inside AI answers, and they are worth it once you are managing more queries than a manual pass can handle. Pricing and features in this category change fast, so treat the following as a map, not a quote — it is based on publicly available pricing and feature listings as of mid-2026, and you should verify current details before buying.

Profound is the most established, enterprise-grade option. It tracks how your brand appears across ChatGPT, Perplexity, and other engines with deep analytics on share of voice, sentiment, and the sources feeding the answers. It does the most, and it is priced for marketing teams and agencies rather than solo operators — the strength is depth, the limitation is cost and complexity for a small local business.

Otterly.AI is more accessible and SMB-friendly. It monitors your chosen prompts and tracks brand mentions and links across AI search surfaces, including Google's AI Overviews, ChatGPT, and Perplexity, on tiered monthly plans. The strength is the gentle on-ramp; the limitation is that lower tiers cap how many prompts and engines you can track, so a query-heavy business outgrows the entry plan.

Goodie is a newer entrant focused specifically on AI brand visibility, positioning itself around helping brands get recommended by AI. The strength is a purpose-built AEO focus; the limitation is that, as a younger product, its coverage and features are still evolving — worth a trial rather than a long contract.

AthenaHQ sits in the generative-engine-optimization category, tracking brand presence across AI engines and pairing measurement with recommendations on what to fix. The strength is that it connects tracking to action; the limitation, again, is that this is a young category with less transparent pricing than mature SEO tools.

General brand-monitoring tools like Brand24 and Mention have started bolting AI-mention tracking onto their broader web-and-social listening. They are convenient if you already use them, but they are built for general brand listening, not citation-rate analytics — for serious AEO measurement they are weaker than the purpose-built options above.

My honest take for most local and small businesses: start free with the manual method and a spreadsheet, prove the channel matters, then add one purpose-built tool once you are tracking more queries than you can check by hand. Do not buy enterprise tooling to monitor ten queries you could test in twenty minutes.

How do you track AEO citations in a spreadsheet?

A spreadsheet is all most businesses need to track citations, and it is free. The key is a consistent structure you fill in the same way every pass, so the data is comparable over time.

Build these columns: query, AI engine, date checked, cited (Y/N), context (the exact sentence), page that should have been cited, and action item. That last pair is what turns tracking into improvement — every "No" should produce a specific fix attached to a specific page.

Here is the setup, step by step.

1. Create the columns above in a blank sheet.
2. Seed it with your 20 queries — one row per query per engine, so 80 rows for four engines.
3. Run the test and fill it in, recording Y/N, the context, and the owning page.
4. Turn gaps into action items — name the exact fix for every "No".
5. Re-check on schedule and date each pass so you can chart citation rate over time.

Check on the same biweekly or monthly cadence as your manual testing, and never delete old passes — the trend is the point. A sheet that shows your citation rate climbing from 15% to 45% over a quarter is the most persuasive AEO report you can put in front of an owner.

What should you do when you're not getting cited?

When you are not getting cited, treat it as a diagnosable problem with a fixed playbook, not a mystery. The engine is naming someone — your job is to find out why it is not you and close the gap.

Work the remediation playbook in order.

1. Find the page that should win the query — or note that it does not exist yet.
2. Add a direct answer and TL;DR so the first sentence answers the question.
3. Add FAQ content and schema so the engine can extract a clean answer.
4. Check your authority signals — see whether the sources it cites instead (Reddit, directories, publications) mention you at all, and start earning presence there.
5. Re-test after the next crawl — wait two to six weeks and re-run the query.

Nine times out of ten, the problem is one of the first three: no clear page, no direct answer, or no schema. Those are fast to fix. Authority is the slow lever, but it is also the most durable once you build it.

What should you do when you ARE getting cited?

When you are getting cited, do not just celebrate — document it and reverse-engineer it, because a win you cannot explain is a win you cannot repeat. Every citation is both proof and a template.

Do three things. First, document the win: screenshot the answer, the query, the engine, and the date. This is real case-study and social-proof material, and it is far more convincing to a prospect than a ranking screenshot. Second, identify which patterns earned the citation — was it the FAQ schema, the direct-answer opening, a Reddit mention, a comparison table? Look at what that page does that your uncited pages do not. Third, replicate across other pages: take the winning pattern and apply it to the next ten queries you want to own. AEO compounds when you treat each citation as a repeatable recipe rather than a lucky break.

How does citation tracking differ by industry?

Citation tracking is the same process everywhere, but what you test and prioritize shifts by industry. Point the same method at the queries that actually drive your revenue.

Local services — dental, gyms, real estate — should weight "best in [city]" and "near me" queries heavily, and test them with location context, because the answer can change based on where the engine thinks you are. A clinic tracking AEO in Kansas City cares about local citation rate far more than national. E-commerce should track product-comparison and "best [product] for [use case]" queries, where being named in the shortlist drives sales. SaaS and B2B should track category and alternatives queries — "best [category] tool," "[competitor] alternatives" — since buyers research with AI long before they book a demo. In every case, the discipline is identical: fixed queries, all four engines, a consistent cadence, and an action item for every gap.

When you are ready to stop checking by hand, a free AEO audit will baseline your citations across every major engine and hand you the gap list. For the strategy behind the fixes, see the complete AEO 2026 guide and AEO vs SEO in 2026.

Frequently asked questions

How do I know if ChatGPT is citing my business? Ask it directly. Run the questions your customers ask — "best [your service] in [your city]", "[your category] near me", and your business name — and note whether you are named, a competitor is, or no source appears. Do the same in Claude, Perplexity, and a Google AI Overview, and repeat every few weeks.

Can I track AI citations for free? Yes. The manual method plus a spreadsheet costs nothing but time and is the right starting point. Paid tools automate and scale it, but prove the channel matters with the free method first.

How often should I check my AI citations? Every two to four weeks for most businesses, with an extra pass after you ship major content or schema changes. A one-off check tells you little; the trend against a fixed query set is what matters.

What is a good AI citation rate? There is no universal benchmark yet, so measure against yourself and your competitors. Watch your share of priority queries trend up, and compare your share of voice to named rivals on the questions that matter most.

Why is my business not showing up in AI answers? Usually because your content is not structured for extraction, your schema is missing, or your authority is thin. The engine cites pages with direct answers, clean schema, and trusted mentions — if a competitor has those and you do not, it names them. Work the remediation playbook above.