Competitors aggregates the brand names that appeared across every answer your tracked prompts returned, and shows how often each one came up.
It answers a question that is hard to ask any other way: in the eyes of the engines, who is the default answer in this market?
Why the list surprises people
The brands the engines name are frequently not the brands you consider your competitors.
Two reasons. Engines answer the question asked, not the question you wish had been asked, so a broad prompt pulls in adjacent categories. And engines favour whoever has been written about most clearly, which rewards companies with strong content over companies with strong products.
Both are worth knowing. The first tells you your prompt was too broad. The second tells you where the gap actually is.
Reading it per prompt
The aggregate is the headline, but the per-prompt breakdown is where the work is. A competitor who appears in every answer is a market leader and not much of an insight. A competitor who owns exactly three of your prompts has done something specific on those three topics, and that is a thing you can go and look at.
Open Sources for those prompts and read the pages being cited.
Turning it into work
A competitor who dominates a topic in AI answers usually also ranks for it in search. Add the terms behind those prompts to keyword tracking to follow your own positions for them, and score your page for the topic with content score, which compares it against the pages ranking in Google for that term. What the competitor gets cited for stays in Sources and Competitors.
That is the most direct path from "they get cited and we do not" to a list of pages to improve.
Track the ones that keep appearing
If a brand you had not considered a competitor shows up in a quarter of your answers, that is the market telling you something. Read the pages they get cited for in Sources before you dismiss it.
What it will not tell you
It does not tell you why a competitor is favoured. The answer is almost always in their content, and reading it is not something this module can do for you.
It also does not measure share of voice in any rigorous sense. Generated answers vary between runs, so treat the counts as a ranking rather than a metric. A competitor at 40% and one at 35% are not meaningfully different. One at 40% and one at 5% are.
Need help?
Write to us and a person answers. See the frequently asked questions first if your question is a common one, or get in touch for anything else.