Direct answer
Yes, more than anything you publish about yourself. A machine treats self-description as an interested claim and discounts it. A mention from a source it already trusts — an independent article, a listing, a review body, a client describing what you did — is the same claim from a party with nothing to gain, which is what makes it repeatable. Volume of reviews matters far less than whether the description is consistent with everything else it reads.
The instinctive objection is that reviews are noisy and mentions are incidental, so surely a well-argued page on your own site counts for more. It counts for less, and the reason is structural rather than editorial. Your page is evidence that you claim something. A third party saying it is evidence that it is true. A system deciding what to repeat in an answer has to distinguish between the two, and it does.
Why self-description is discounted by default
Everything on your own domain is written by an interested party. That is not an accusation; it is a fact about the source, and any system weighing evidence applies it. So the page announcing that you are the leading firm in your niche is read as an assertion, filed, and largely set aside. The same sentence in an independent article is read as a finding. Nothing about the wording changed. The party saying it did.
You can publish a claim. You cannot publish a confirmation. That asymmetry is the whole game.
Not all mentions weigh the same
The useful sort names you, places you in a category, and comes from somewhere with its own reputation to protect. The useless sort is a link in a directory nobody reads, or a paid placement whose pattern is obvious. Between them sits the bulk of what firms actually collect — real but vague mentions that name you without saying what you do, which confirm existence and nothing else.
| Type of mention | What it confirms | Weight |
|---|---|---|
| Independent article describing your work in your category | You exist, in this category, doing this | High |
| Named client or reviewer describing a specific outcome | The claim held up in a real case | High |
| Listing or profile with a consistent description | Your identity is stable across sources | Moderate |
| Volume of short, generic five-star reviews | Little — the pattern is cheap to produce | Low |
| Paid placement in an obvious network | That someone paid | None or negative |
Reviews: the count is the least interesting number
Firms chase review volume because it is countable. But a hundred reviews saying “great service” describe nothing a machine can attach to a category or a claim. Five reviews that say what the work was, in what sector, with what result, do more, because they contain the facts that let a system place you. Specific beats numerous, and it isn’t close.
Working benchmark: across the diagnostics I’ve run, corroboration is the weakest of the four citability signals more often than any other — firms typically have an identity and a site, and almost nothing outside their own control describing them in their category. It is also the slowest to repair, which is why it’s worth starting before it is urgent. An internal pattern from client work, not a published statistic.
What actually earns them
Being quotable and being available. A firm that publishes a specific, checkable claim gives journalists, analysts and clients something to repeat; a firm that publishes adjectives gives them nothing. Then it is unglamorous work: answering questions publicly, letting client results be named, being the source someone cites because you said something precise. That is the difference between authority and spend, and it settles the question of whether authority building is worth it or a vanity cost.
Why this is the signal competitors beat you with
A firm with a weaker product and three good independent mentions gets named ahead of a better firm with none, because the machine can verify one and not the other. That is uncomfortable and it is also the mechanism behind why buyers shortlist competitors you are objectively better than. The remedy is not louder self-description. It is getting one credible outsider to say the thing you have been saying alone.
- Search your category and list every source describing you that you don’t control. If the list is short, that’s the gap.
- Check that those sources describe you the same way — same category, same specialism.
- Ask two clients for a review that names the work and the result, not the feeling.
- Publish one specific, checkable claim a journalist or analyst could quote.
- Stop counting reviews. Count how many say something a machine could place you by.
If you want to see what outside sources currently say about you, and what an engine does with it:
Keep writing your own case and nobody else’s, and the machine will keep filing you as a company that says it is good — next to one that someone else said it about.
Book a free 15-minute session See the proof