← Blog
#ai-visibility#competitors#geo

How Do I Know If AI Is Sending Customers to My Competitors?

September 10, 2026

How Do I Know If AI Is Sending Customers to My Competitors?

Ask an assistant the questions your buyers ask before they know your name, and record who gets named. If competitors appear consistently and you do not, that is the answer. You do not need traffic data to see it, which is fortunate, because traffic data will not show it clearly anyway.

Table of Contents

Why your analytics will not tell you

Referrals from assistants are messy. Some arrive with an identifiable referrer, many arrive as direct traffic after someone reads an answer and types your name, and a buyer who never saw your name leaves no trace at all. The most expensive version of this problem, being absent from a shortlist, is by definition invisible in your analytics.

So the measurement has to happen on the answer side, not the traffic side.

The three signals to look for

1. Consistent competitor naming. Run each buyer question several times. If the same two competitors appear in most runs and you appear in none, you are not unlucky, you are absent.

2. The cited sources. Look at which domains the answer credits. Those are the pages doing the work. If a comparison article names four vendors and not you, that single page is likely costing you more shortlist appearances than your entire blog is earning.

3. Stage asymmetry. Check discovery, evaluation and decision questions separately. Most brands are not uniformly absent. In our own tracked data, companies show up in evaluation-stage answers considerably more often than in discovery or decision answers, which means the losses are usually concentrated at one end of the funnel rather than spread evenly.

What to do with the answer

  • Take the list of cited domains and treat it as an outreach and inclusion list. Getting added to an article that is already being quoted is usually faster than trying to get a new page quoted.
  • Publish the honest head-to-head against the competitor who keeps appearing. Assistants pull comparison-shaped content heavily for shortlist questions.
  • Fix consistency. If your description of what you do varies across your site, directories and coverage, there is no stable claim for a model to repeat.
  • Then measure again on the same prompts, so you know whether it worked rather than guessing.

The free check gives you the competitor picture on a sample of buyer prompts with no signup, and how Monroya works covers the loop after that.

FAQ

Can I see how much revenue this costs me? Not precisely, and be careful with anyone who claims otherwise. What you can measure honestly is shortlist presence, and how it changes when you act.

Are AI referrals even meaningful yet in B2B? Volume is smaller than search, but the intent is late-stage: someone asking which vendor to pick is close to a decision. Presence on those questions is worth more per appearance than a discovery impression.

What if a competitor is much larger than us? Size helps them, but it is not decisive here. Assistants pull from sources, and a specific, well-structured comparison page can outrank a large company's generic page as a citation. That is why smaller competitors sometimes outrank larger companies.