“Are we visible in ChatGPT?” is the wrong first question. The better one is: “who is AI recommending instead of us, and why?”

A visibility score in isolation is close to meaningless. If your product shows up in 20% of relevant prompts, is that good or bad? You cannot tell without knowing what your direct competitors score on the same prompts. This guide walks through a practical way to track competitor visibility alongside your own.

Why competitor context changes everything

Say your SaaS appears in 3 out of 10 ChatGPT answers about your product category. On its own, that number tells you almost nothing. But if your closest competitor appears in 8 out of 10 on the exact same prompts, you now know two things: there is a real gap, and it is closable, because the category is clearly being discussed and recommended at all.

The reverse is also useful. If nobody in your category - including competitors - shows up reliably, that tells you the category itself is not yet well represented in AI training data or retrieval sources, which is a different problem with a different fix.

Step 1: Pick the right competitors

Do not track every company that competes for the same keywords in Google. Track the 3-5 products a prospective buyer would realistically compare you against right before signing up. For most indie SaaS founders, this means:

  • Direct feature-for-feature competitors in your category
  • The 1-2 category leaders, even if they are bigger than you
  • Any product that consistently comes up in your own sales calls or churn interviews

Step 2: Run the same prompts for everyone

This is the part that is easy to get wrong. Comparing your visibility on “best CRM for small teams” against a competitor’s visibility on “top CRM software” is not a fair test - different prompts surface different answers.

Run an identical prompt set across every brand you are tracking. This is exactly what AskRank’s competitor comparison does automatically: the same prompt library, the same LLMs, the same schedule, for you and up to 5 competitors at once.

Step 3: Track position, not just presence

Being mentioned fourth in a five-product list is very different from being mentioned first. Two brands can both have a 50% mention rate and mean completely different things for revenue: one is consistently the top pick, the other is a footnote.

Track three dimensions per competitor, per prompt:

  1. Mentioned or not - the basic binary signal
  2. Position - where in the answer they appear relative to you
  3. Sentiment - whether the AI frames them positively, neutrally, or with caveats

Step 4: Find the gap prompts

The most actionable output of competitor tracking is not the overall score. It is the specific list of prompts where a competitor is mentioned and you are not. These are called gap prompts, and they point directly at where to focus content and citation-building work.

If a competitor consistently wins on “best tool for solo founders” and you do not, that tells you exactly what page to write next, and which positioning to sharpen on your own site.

Step 5: Watch the trend, not a single snapshot

LLM answers are not fully deterministic. The same prompt can return slightly different results run to run. A single comparison is a data point, not a trend. Track competitor visibility on a recurring schedule and look at the trend line over weeks, not the result of one manual check.

This is also why multi-sampling matters: running each prompt more than once per cycle smooths out noise so a real shift in visibility is not lost in normal variance.

What to do once you see the gap

Once you know where competitors are winning, the fix usually falls into one of three buckets:

  • Content gap - write the page that directly answers the prompt where you are losing
  • Citation gap - get mentioned in the third-party sources (G2, Capterra, comparison blogs) that the AI is pulling from
  • Positioning gap - your own site is not clear enough about who you are for, so the AI cannot confidently recommend you

For a deeper look at how citation-driven visibility works, see our guide on answer engine optimization.

Start tracking today

Manually running the same 10-20 prompts against yourself and 3-5 competitors, across four LLMs, on a recurring schedule, is not something you can sustain by hand for more than a week or two. AskRank automates the entire loop, and the free tier includes one competitor comparison to start with.