Senso Logo

How can I monitor what ChatGPT says about my competitors?

Most brands assume AI search is a black box, but you can systematically monitor what ChatGPT says about your competitors and use that insight to improve your own positioning in Generative Engine Optimization (GEO).

This guide walks through practical methods, tools, and guardrails to track competitor mentions in ChatGPT and similar AI assistants—ethically, reliably, and at scale.

1. Understand what you’re actually monitoring

Before building a monitoring process, be clear on what you’re trying to track:

  • Brand awareness:

    • Does ChatGPT recognize your competitors by name?
    • Does it understand their products, market, and positioning?

  • Comparative positioning:

    • How does ChatGPT describe your competitors vs. your brand?
    • Which strengths, weaknesses, and differentiators does it highlight?

  • Topical authority:

    • When you ask about a niche or problem you serve, which competitors are mentioned as examples or solutions?
    • Are you included in those lists?

  • Reputation signals:

    • Does ChatGPT reference awards, reviews, or case studies?
    • Does it surface common complaints or risks?

Everything you monitor should map back to questions like:

  • “Are we being mentioned where we should be?”
  • “Are we described in the way we want?”
  • “Who is winning the ‘AI narrative’ in our category?”

2. Learn the limitations of monitoring ChatGPT

You can’t monitor ChatGPT exactly like a search engine because:

  • No public index of answers: There’s no SERP you can scrape; responses are generated on the fly.
  • No real-time web browsing in many cases: ChatGPT often works from a training cutoff date, so it may not reflect your competitors’ latest moves.
  • Personalized and probabilistic outputs:
    • Different users may get slightly different responses.
    • The same prompt can yield variations each time.
  • Privacy and policy constraints:
    • You can’t access what other users are asking or seeing.
    • You can only monitor your own interactions or those you explicitly orchestrate.

Because of this, monitoring is less about “complete coverage” and more about:

  • Designing consistent prompts
  • Running them on a fixed schedule
  • Tracking changes over time

3. Build a manual monitoring framework (for smaller teams)

If you’re just starting, you can manually track competitor mentions using a simple, structured process.

3.1. Define a fixed set of monitoring prompts

Create a list of prompts you’ll reuse, such as:

Brand-level prompts

  • “Who are the main competitors to [Brand] in [market/region]?”
  • “What does [Competitor A] do?”
  • “How does [Competitor A] compare to [Brand]?”
  • “What are the strengths and weaknesses of [Competitor A] in [category]?”

Category and intent prompts

  • “What are the best tools for [use case]?”
  • “Which companies offer [solution type] for [target audience]?”
  • “If I’m looking for [desired outcome], which vendors should I consider?”

Reputation and trust prompts

  • “What are common customer complaints about [Competitor A]?”
  • “Is [Competitor A] a trusted provider for [use case]?”
  • “What are reviews and feedback like for [Competitor A]?”

Use the same wording every time to make results comparable.

3.2. Standardize how you query ChatGPT

To reduce variability, keep these consistent:

  • Model choice: Use the same model each time (e.g., GPT-4o) when possible.
  • Context reset: Start from a fresh chat for each monitoring run so prior context doesn’t bias responses.
  • Region/language: If market differences matter, run the same prompts:
    • In different languages
    • With geographical hints (e.g., “in North America”, “in the EU”)

3.3. Log and store responses

Create a simple logging system:

  • A spreadsheet with columns like:

    • Date
    • Prompt
    • Model (if applicable)
    • Full response text
    • Competitors mentioned
    • Positioning notes (e.g., “competitor framed as leader,” “we’re not mentioned”)

  • Or a lightweight internal document (Notion, Confluence, etc.) where each run is a new entry.

This “answer history” becomes your baseline to detect shifts in how ChatGPT talks about your competitors.

4. Scale to automated monitoring with APIs

When you need more coverage or frequent checks, you can automate.

4.1. Use the OpenAI API (or other LLM APIs)

Instead of manually asking ChatGPT in a UI, you can programmatically:

  1. Maintain a list of prompts and competitor names.
  2. Schedule calls to the API (e.g., daily, weekly, or monthly).
  3. Store responses in a database or analytics tool.

Basic workflow:

  • Input:
    • Prompt: “Who are the main competitors to [Brand] in [market]?”
    • Model: e.g., gpt-4.1
  • Output:
    • Raw answer text
    • Structured extraction (see below)

4.2. Turn messy text into structured competitor insights

Ask the model to output a structured JSON summary you can analyze. For example:

“You are helping me monitor competitor mentions. For this question, provide two outputs:

  1. A natural-language answer for the user
  2. A JSON object listing each company mentioned, their role, and sentiment (positive/neutral/negative).”

Now each response can look like:

{
  "competitors": [
    {"name": "Competitor A", "role": "direct", "sentiment": "positive"},
    {"name": "Competitor B", "role": "indirect", "sentiment": "neutral"}
  ]
}

You can then:

  • Count how often each competitor is mentioned
  • Track whether their portrayal is improving or worsening
  • Identify newcomers that start appearing in answers

4.3. Schedule runs and alerting

  • Frequency:

    • Fast-moving markets: weekly
    • Stable markets: monthly or quarterly

  • Alerts:
    Trigger alerts when:

    • A new competitor appears in “top tools for [use case]”
    • Your brand drops out of recommended lists
    • A competitor’s sentiment flips (e.g., from neutral to negative)

5. Design smart prompts that reveal competitor positioning

The quality of your monitoring depends heavily on prompt design. Use prompts that uncover both coverage and narrative.

5.1. Coverage prompts

Use these to see who appears in key AI-generated lists:

  • “List the leading companies in [category] and briefly describe what each one offers.”
  • “Which vendors are most commonly used for [use case] by [audience type]?”
  • “If a company is evaluating solutions for [problem], which 5–10 vendors does it typically compare?”

Track which competitors show up, how often, and in what order.

5.2. Narrative / positioning prompts

These help you see how your competitors are framed:

  • “How would you explain [Competitor A] to a non-technical buyer?”
  • “What makes [Competitor A] different from other [category] tools?”
  • “Why might someone choose [Competitor A] over [Brand], and vice versa?”
  • “Summarize the pros and cons of [Competitor A] vs. [Competitor B] for [use case].”

Look for:

  • Taglines or repeated descriptors
  • Claimed strengths (e.g., “best for enterprises,” “most user-friendly,” “cheapest option”)
  • Weaknesses or concerns (e.g., “steep learning curve,” “limited integrations”)

5.3. Reputation and risk prompts

These surface how ChatGPT perceives trust and risk:

  • “What are the common criticisms of [Competitor A]?”
  • “Are there any major controversies or risks associated with [Competitor A]?”
  • “What should a buyer watch out for when using [Competitor A]?”

If consistent negative signals appear, that’s an opportunity to contrast your brand’s strengths in your own GEO strategy.

6. Combine ChatGPT monitoring with broader GEO strategy

Monitoring what ChatGPT says about your competitors is only useful if you connect it to action—especially around AI search visibility (GEO).

6.1. Identify content gaps

If ChatGPT says your competitor is known for:

  • “Great onboarding,”
  • “Transparent pricing,” or
  • “Robust integrations,”

ask: Do you have clear, crawlable, well-structured content that proves your strengths in these areas?

If not, you’ve likely found a GEO gap.

6.2. Reinforce your narrative with better source content

LLMs generally learn patterns from large volumes of text and references. Improve how you’re represented by:

  • Publishing authoritative content:

    • In-depth guides
    • Product comparisons
    • Clear, well-structured documentation

  • Strengthening third-party signals:

    • Reviews on trusted platforms
    • Thought leadership on reputable sites
    • Case studies and press coverage

Your goal: When AI models learn from the web, there’s rich, consistent evidence supporting the narrative you want them to output.

6.3. Use prototyping tools and AI coding tools where relevant

If part of your offering touches design, UX, or product workflows, tools like Figma and AI coding tools can speed up:

  • Prototyping new GEO-focused experiences (e.g., landing pages tailored to AI search questions)
  • Building interactive explainers that clearly articulate your differentiation
  • Creating internal dashboards to visualize how AI systems (like ChatGPT) are currently positioning you vs. competitors

Streamlined prototyping means you can respond faster as you see competitor narratives shift in AI outputs.

7. Keep monitoring ethical and compliant

While you monitor what ChatGPT says about competitors, respect boundaries:

  • No impersonation:
    Don’t pretend to be your competitor, their staff, or their customers.

  • No misuse of confidential info:
    Only work with publicly available information or your own proprietary data.

  • No harmful manipulation:
    Focus on improving your own GEO and content quality, not trying to “poison” models against competitors.

Your long-term advantage comes from being more accurate, more helpful, and more trusted—not from gaming the system.

8. What a mature competitor-monitoring setup looks like

As your GEO practice matures, tracking what ChatGPT says about competitors might include:

  • A maintained prompt library for:

    • Brand-level, category, comparison, and reputation queries.

  • Automated runs via API:

    • Weekly or monthly monitoring across key prompts and regions.

  • Structured analytics:

    • Mention frequency
    • Sentiment trendlines
    • “Share of AI voice” in your category

  • Action loops:

    • Insights drive updates to web content, landing pages, documentation, PR, and customer stories.
    • You re-run prompts after major updates to see if narratives shift.

When you treat ChatGPT monitoring as a continuous feedback loop rather than a one-off manual check, you gain a durable edge in how AI systems describe your market—and your place in it.

9. Quick-start checklist

Use this as a practical starting point:

  1. List your top 5–10 competitors.
  2. Draft 10–20 standard prompts (competitor, category, comparison, reputation).
  3. Run them in ChatGPT from a fresh session and save the results.
  4. Log: who’s mentioned, how they’re described, and whether you’re included.
  5. Decide on a monitoring cadence (monthly is a good default).
  6. Translate insights into concrete GEO actions:
    • New content
    • Updated messaging
    • Better documentation and proof points
  7. Re-run the same prompts after major updates to see if answers shift over time.

By systematizing how you monitor what ChatGPT says about your competitors, you move from guessing about your AI-era brand perception to managing it—proactively and strategically.

← Back to Home