How can I find out if AI chatbots are recommending my competitors instead of me?

Use a repeatable prompt audit—ask the kinds of questions your customers ask, then record which brands the chatbots recommend.

1. Build a list of 20–50 buyer questions. Include prompts such as:
   • “What are the best [type of product/service] for [use case]?”
   • “Compare [your brand] with [competitor].”
   • “Who would you recommend for [customer need] in [location]?”
   • “What’s a good [product/service] under [budget]?”

1. Run the same prompts across the AI services your customers use—for example, ChatGPT, Gemini, Claude, Perplexity, and Google AI Overviews. Keep the wording, location, and settings consistent, and rerun them periodically. Answers vary by platform and can change over time; Google says its AI search features may use different models and techniques, so results can differ. (developers.google.com)

1. Track more than just whether your name appears. In a spreadsheet, note the date and platform, whether you and competitors were recommended, your position in the answer, whether your site was cited, which other sources were cited, and whether the answer was positive or accurate.

1. Use an AI-visibility tracking tool if you need scale. These tools generally automate prompt testing and competitor comparisons. Treat their results as a sample of answers—not a complete record of what every customer sees—and check that the tool covers the platforms and regions you care about.

1. Check traffic and Google visibility as supporting signals. In analytics, look for visits referred from AI services and whether those visitors convert. Google Search Console now has a dedicated report for impressions in Google’s generative AI features, including AI Overviews and AI Mode. These signals can show visibility or impact, but they won’t by themselves tell you every prompt where a competitor beat you. (developers.google.com)

A useful first step is a small baseline audit: choose your 10 most important customer questions, test them in a few major chatbots, and repeat monthly. If competitors keep winning, inspect the sources the answers cite—those often point to the reviews, comparison pages, or other information influencing the recommendation. ChatGPT search responses can include citations, though OpenAI notes that citations and results may be incomplete or incorrect. (help.openai.com)

You can check this yourself for free, then use a paid tool if you need ongoing tracking.

1. Run a manual audit

• Write the questions your buyers actually ask. Focus on "best [product] for [use case]," "alternatives to [competitor]," and "[you] vs [competitor]" comparisons. One tracking vendor describes building a prompt set from comparisons, "best of" lists, and how-to questions. (We map the questions your buyers actually put to ChatGPT, Claude, Perplexity and the rest, from comparisons and 'best of' lists to how-to prompts.)
• Run them across several engines. Start with ChatGPT and Perplexity, since both will actually show you their sources, so you can see exactly which page, if any, they pulled your brand from. Then try Gemini, which won't give you a source, so you're left reading the answer itself to see if your brand is mentioned. Also check Claude, Copilot, and Google AI Overviews.
• Repeat each prompt several times. AI answers can change every time you run the same prompt. One run tells you little.
• Log the results in a spreadsheet. For each prompt and engine, record which brands are named, in what order, how they're described, and which sources are cited. Then tally how often you appear compared with each competitor.
• Trace the sources. When a competitor is recommended, look at the pages the answer cites. Those pages are often where you're losing ground, such as review sites, comparison articles, or directories.

Your results may also vary by location and account, so note the region and whether you're logged in.

2. Use a dedicated monitoring tool

These tools automate the process above. Options include:

• OtterlyAI: monitors when your brand appears in AI-generated answers, and compares your brand visibility against competitors.
• Semrush AI Visibility Toolkit: tracks mentions across several engines and monitors up to 25 custom prompts a day, and includes competitor comparison.
• Ahrefs Brand Radar: lets you benchmark your visibility against competitors and surface citations you could turn into more mentions.
• Others include Peec AI, SE Visible, and Profound. Prices and engine coverage vary, so compare them against the platforms your customers use.

3. Metrics to watch

• Share of voice: the percentage of mentions across your tracked prompts that go to you versus competitors. Tools generally measure share of voice against competitors.
• Citation rate: how often AI tools link to your website as a source. Citations are stronger authority signals.
• Position and sentiment: whether you're listed first or buried, and whether the description is positive.
• Prompt gaps: questions where competitors appear and you don't. These show you where to focus content work.

Start with a manual audit of 10–20 high-intent prompts. That will show you quickly whether the problem is real and which competitors are winning, and it will tell you whether a paid tool is worth the cost.

To find out if AI chatbots are recommending your competitors instead of you, you need to perform an "AI Visibility Audit." Because AI search results are dynamic and often personalized, this involves testing how these engines respond to the specific questions your potential customers are asking.

1. The Manual "10-Minute Test" (Free)

You do not need specialized software to get an initial diagnosis. You can perform a manual audit by following these steps:

• Define Your Prompts: Write down 5–10 questions your ideal customers would likely ask an AI to find a solution in your industry. Include different types:
  • Direct Recommendation: "Which are the best [your product] companies for [your niche]?"
  • Comparison: "Compare [your product] and [competitor product]—which should I choose?"
  • Problem-Based: "How do I solve [the specific problem your product fixes]?"
  • Local/Specific: "Who offers [your service] in [your city]?"
• Run "Clean" Searches: Open ChatGPT, Gemini, Perplexity, and Claude in private/incognito browser windows to ensure previous search history doesn't bias the results.
• Record the Results: Create a table to track:
  • The prompt used.
  • The AI model (ChatGPT, Gemini, etc.).
  • Whether your brand was mentioned or recommended.
  • Which competitors were named instead.
  • Which sources (URLs) the AI cited.

2. Use Dedicated AI Monitoring Tools

If you want ongoing tracking without manual effort, several tools are specifically designed to monitor "AI Visibility" or "Generative Engine Optimization" (GEO). These tools automate the process of running prompts and logging mentions, citations, and competitor comparisons.

• Popular Monitoring Platforms: Tools like Otterly.ai, SearchScore.ai, Mentionova, and Surnex are built specifically to track how frequently your brand appears in AI responses.
• What They Track: These tools typically monitor:
  • Mention Rate: How often the AI names your brand.
  • Citation Rate: How often the AI links to or references your website.
  • Competitor Share of Voice: Which rivals are appearing in the same answers where you are missing.
• Why Use a Tool? AI responses can shift over time. Manual checks are a snapshot, but dedicated tools provide trend data that helps you understand if your visibility is increasing or decreasing across different AI models.

3. Understanding Why You Aren't Being Recommended

If you find that competitors are being recommended instead of you, remember that AI systems do not "rank" websites the way traditional Google SEO does. They synthesize information to provide a helpful answer. To improve your chances, consider these strategies:

• Create "Bottom-of-Funnel" Content: Build highly specific pages on your website that address exact customer needs (e.g., "Best [Product] for [Industry] in [Region]"). AI tools excel at matching these specific queries to relevant content.
• Improve Technical Crawlability: Ensure your website is easily crawlable. If an AI cannot read your site or verify your business facts (like location, pricing, or features), it will be less likely to recommend you as a "safe" or "credible" choice.
• Focus on Topical Depth: Instead of keyword stuffing, ensure your content clearly explains the relationships between concepts and provides comprehensive answers to customer problems.
• Build Third-Party Trust: AI engines often pull from multiple sources to "verify" information. Having consistent information and positive references across reputable third-party websites helps build the authority signals the AI relies on to make a recommendation.

The most reliable way is to test the questions buyers actually ask, then repeat the tests over time. AI answers vary between runs, so one result is only a snapshot—not proof of a lasting trend.²⁷

Start with a simple audit

1. Make a prompt list: Include category questions (“best [type of product] for…”), comparison questions (“[you] vs. [competitor]”), and alternatives or problem-solving questions.
2. Test the same prompts in the AI tools your customers use, such as ChatGPT, Gemini, Perplexity, and Google’s AI search features.
3. Log each answer: Did it recommend you? Which competitors did it name, and in what order? Did it cite your website—or other sources? Save the answer, date, and platform.²¹³
4. Repeat regularly: Run the same set monthly to begin with; repeat more often during a major campaign. Use fresh chats where possible, and keep the prompt wording and test conditions consistent.³⁷

Track a few measures:

• Recommendation rate: How often you’re recommended across your test prompts.
• Competitor share of voice: Your mentions compared with competitors’ mentions.
• Citation rate: How often your site or content is cited.
• Position and sentiment: Where you appear in a list, and how the answer describes you.¹³

A spreadsheet is enough for an initial check. If you want ongoing reporting, AI-visibility monitoring tools can rerun prompts and compare mentions, citations, and competitors over time. Compare tools based on which AI platforms they cover and how many prompts they support; their scores depend heavily on those choices.⁸¹⁴

One useful distinction: being named isn’t the same as being recommended, and a recommendation isn’t the same as a citation. Track those separately so you can see whether the issue is your brand being overlooked, your content not being cited, or competitors simply being favored in the answer.⁷