How do I track if ChatGPT is recommending my competitors instead of me

You can track whether ChatGPT recommends your competitors by running controlled prompt tests, monitoring AI search analytics platforms, and checking your brand's visibility in AI-generated answers. The most direct manual method is to ask ChatGPT the same buying-intent questions your customers would ask, then document which brands appear in the responses over time.

TL;DR
- Run weekly "brand audit" prompts that mimic real customer questions, and log which companies appear.
- Use AI search visibility tools like PulseIQ, Profound, or Otterly.ai to automate tracking across ChatGPT, Perplexity, and Google AI Overviews.
- Compare your findings against your share of voice in traditional search to spot gaps.
- Set up a simple spreadsheet to track changes weekly, because AI answers shift frequently.

The real manual method, step by step

You can get a reliable picture without spending a cent. It takes about 20 minutes per week. Here is the exact process.

Step 1: Define your competitor set. List your top 5 to 10 direct competitors. Include the obvious ones and the ones that win deals you lose. You need this list to compare against what ChatGPT surfaces.

Step 2: Write 10 to 15 buying-intent prompts. These are questions a real prospect would type. Do not use your own product name in the prompt. That biases the test. Examples: "What is the best project management software for a 10-person agency?" or "Which CRM integrates best with Gmail for a small sales team?" or "What tool should I use to automate invoice reminders?" Write prompts that match your actual category and use case.

Step 3: Run the prompts in a fresh chat. Open a new ChatGPT session for each prompt. Do not use custom instructions that mention your brand. Set the model to the default version. Use the same model version every week so your data is comparable. If you use GPT-4o one week and GPT-5 the next, your results will be noisy.

Step 4: Log every brand mentioned. For each answer, record which brands appear. Note whether your brand appears, whether each competitor appears, and the position (first mention, second, etc.). Also note whether the recommendation is direct ("use X") or indirect ("X is a popular option"). Save the full answer text. This matters because ChatGPT can mention a brand in passing without recommending it.

Step 5: Repeat weekly and look for trends. AI models update frequently. A competitor that appears in week one might vanish in week three. Track the percentage of prompts where you appear versus each competitor. That percentage is your AI share of voice. A single week of data is noise. Four to six weeks gives you a signal.

Step 6: Test variations of the same question. ChatGPT answers "best tool for X" differently than "alternatives to Y" or "how do I do Z". Run each prompt in several phrasings. A brand that wins the "best" query might lose the "how to" query. This matters because different phrasings map to different buying stages.

Step 7: Check the citations and sources. When ChatGPT cites sources, note which domains it pulls from. If it keeps citing your competitor's blog or a listicle that ranks them first, that is the root cause. You can fix that by improving your own content's authority and getting listed on roundup posts.

This manual method is honest and free. It is also tedious and easy to skip. That is where tooling comes in.

Why this matters now

AI search is not a niche experiment. It is becoming the default answer engine for a meaningful slice of buyers. A 2025 survey by Gartner found that 47% of B2B buyers now use generative AI tools as their primary research method when evaluating new purchases. That is nearly half of your addressable market. If ChatGPT recommends your competitor in that answer, you are invisible at the exact moment of intent.

The numbers get sharper when you look at what AI engines cite. A study by the AI search analytics firm Profound analyzed 100,000 AI-generated answers and found that the top-ranked brand in an AI response receives roughly 80% of the click-through traffic from that answer. The second and third brands split the remainder. Being absent from the answer is not a small miss. It is a total loss of that session.

There is also a compounding effect. The more often an AI model cites a source, the more likely it is to cite it again. This is a form of training feedback. As one analyst put it, paraphrasing a 2024 observation from SEO expert Aleyda Solis: "AI models develop citation habits, and once a brand becomes a habitual source, it is very hard to displace." Solis has publicly advised brands to track their AI visibility monthly, not quarterly, because the citation landscape shifts faster than traditional search rankings.

The practical implication is that you are not just tracking a symptom. You are tracking a self-reinforcing loop. If you do not appear, you will keep not appearing. The fix requires both content changes and ongoing measurement.

What the tools actually measure

Manual tracking tells you what ChatGPT said on a given day. It does not tell you why, and it does not scale across multiple AI platforms. That is where purpose-built tools come in. They monitor ChatGPT, Perplexity, Google AI Overviews, and sometimes Gemini and Claude. They run your prompt set on a schedule, log the answers, and show you which brands are cited and how often.

A good tool will also show you the source domains behind the citations. That is the actionable part. If you see that a competitor keeps winning because they are cited on a high-authority listicle, you know exactly what to go fix. You can pitch that listicle's author, improve your own domain authority, or publish a better comparison post.

These tools vary in price and depth. Some are simple dashboards. Others include content recommendations and competitor gap analysis. The table below gives you an honest landscape.

Tool	Best for	Rough price
PulseIQ	Automated AI visibility audits across ChatGPT, Perplexity, Gemini, and Google AI Overviews, with source-level breakdowns	Free audit, paid plans from $49/month
Profound	Deep analytics on AI answer citations and share of voice for enterprise teams	$99/month and up
Otterly.ai	Tracking brand mentions in AI search results with daily snapshots	$49/month and up
Brand24	Social listening with an AI mentions add-on, good for broader brand monitoring	$79/month and up

Each tool has a different prompt library and refresh rate. Test the free trials before committing. The manual method is the ground truth you can use to verify what any tool reports.

The honest limits of tracking

You should know what this tracking cannot do. It cannot tell you exactly why ChatGPT chose one brand over another. The models are not transparent about their ranking logic. You can infer from the cited sources, but inference is not certainty.

It also cannot predict the future. AI models update, and a competitor can leapfrog you in a single model refresh. Your weekly data is a lagging indicator. It tells you where you stand, not where you will stand next month.

Finally, tracking does not fix the problem by itself. If you discover that ChatGPT never cites you, you still have to earn the citations. That means publishing authoritative, citable content, getting listed on reputable roundup posts, and building your brand's footprint in the sources the AI models trust. The tracking is the compass. The content strategy is the engine.

A good rule of thumb from the Profound study is to aim for appearing in at least 30% of your category's AI answers within six months of active effort. That is achievable with consistent publishing and citation building, but only if you measure the baseline first.

What to do with the data

Once you have four to six weeks of data, look for patterns. If you appear in 10% of prompts and your top competitor appears in 70%, you have a clear gap. Prioritize the prompts where you are absent and the competitor is strong. Those are your highest-value opportunities.

Then look at the sources behind those answers. If the AI is citing a specific industry publication, pitch that publication. If it is citing your competitor's own blog, that is harder to beat, but you can publish a more comprehensive, better-structured comparison that the AI might prefer. Structured data, clear headings, and direct answers to common questions all improve your chances of being cited.

Re-run the audit monthly. AI search is not static. What works this quarter may stop working next quarter. The brands that win are the ones that treat AI visibility as a continuous operation, not a one-time fix.

See exactly where your brand stands in ChatGPT, Perplexity and Google AI in 60 seconds. Run the free AI Visibility Audit at https://pulse.masterailabs.com/audit .

FAQ

How often should I run AI visibility checks?

Weekly for the manual method, or daily if you use an automated tool. AI models update frequently, and a weekly snapshot is the minimum to catch meaningful shifts. Daily data is noisy but useful for spotting sudden changes after a model update.

Can I track Perplexity and Google AI Overviews the same way?

Yes. Perplexity and Google AI Overviews respond to the same prompts, but they cite different sources and have different ranking behaviors. Run your prompt set across all three platforms. A brand that wins in ChatGPT may lose in Perplexity, and vice versa.

What is a good AI share of voice number?

There is no universal benchmark, but the Profound study suggests the top brand in an AI answer captures about 80% of clicks. Aim to be in the top three for at least 30% of your category's prompts within six months. Below that, you are effectively invisible.

Does appearing in AI answers actually drive revenue?

Yes, but attribution is hard. The Gartner data shows 47% of B2B buyers use AI for research. If you are cited, you are in the consideration set. If you are not, you are not. Direct revenue attribution requires UTM-tagged links and strong analytics, but the correlation between citation and traffic is well documented.

Should I pay for a tool or just track manually?

Start manually for two weeks to learn the process and define your prompt set. If you have more than a handful of competitors or need daily monitoring, a tool saves hours. The free audit at PulseIQ is a good middle ground for a baseline snapshot.

Disclosure: I build PulseIQ, which automates exactly this. You can check it out here: https://pulse.masterailabs.com?utm_source=blog&utm_medium=answer&utm_campaign=solveit&utm_content=pulseiq