Mentioned, and still weak

Meta description: A weak brand signal is a public record with no clear reason to remember the company — fixable, but only at the source.

The brand team ran the check and felt reassured: the models know the company. Asked directly, each engine produces a description — accurate name, right industry, reasonable summary. The reassurance lasted until someone asked the buyer’s question instead of the brand’s: not “what is this company” but “who should I consider for this problem” — and the company either did not appear, or appeared in a list with a description so generic that no buyer could act on it. The brand had presence. Presence, I have learned to explain, is a low bar. The models knew the company the way one knows a name from a directory: filed, findable, and saying nothing.

[AWAITING MIKLOS — scene: the model can name the company and cannot say anything distinctive that a buyer would act on]

What can the model say about the brand today?

This is the test I run first, and I recommend its discomfort: ask the engines, in the buyer’s phrasing, and read what comes back as a buyer would. In the companies I am describing, the descriptions share a shape: accurate category, vague capability, no distinction. “A provider of solutions in the X space” — the sentence that fits four hundred companies. The description is not wrong; wrong would be fixable. It is empty, and empty is a different diagnosis. The model has enough signal to place the company and not enough to say anything about it. The brand exists in the model the way a stub exists in an encyclopedia: an entry, awaiting content.

Why is that different from being unmentioned?

What I count as a signal: brand familiarity; a consistent footprint of external discussion in publications and professional communities; content distinctive enough that a reader — human or machine — can say what this company does and for whom. A weak signal is not a small one; it is an unclear one. If the public record gives the answer layer nothing specific to repeat, the company is not being rejected. It is being forgotten at scale.

The distinction matters because the remedies differ. Unmentioned is a presence problem: the company is absent, the task is to become known. Weak signal is a description problem: the company is known, and what is known is unusable. I have watched companies misdiagnose the second as the first and respond with presence-building — more mentions, more coverage, more volume. The mentions accumulate and the description stays generic, because the additions carry the same emptiness as the existing record. The company is buying more of the thing that is not working. Unmentioned brands need volume. Weak-signal brands need something volume cannot supply, and the confusion between the two conditions is expensive.

Why do mention campaigns leave the description generic?

Read more: more about Miklos Roth.

The false explanation is reach: more placements, more press, more citations across more sites. I have traced the record behind generic descriptions, and the problem is not quantity — it is that every mention says the same nothing. Press releases repeating boilerplate. Directory listings with the category field and the founding year. Guest articles that describe the market, not the company. The record is extensive and information-free: a thousand sources agreeing that the company exists and is in the industry. Models compress what the record contains. A record containing nothing distinctive compresses to a description containing nothing distinctive, at whatever volume.

What pattern marks a weak signal?

The marker I use is the substitution test: take the model’s description of the company and substitute a competitor’s name. If the description still works — if it describes the competitor equally well — the signal is weak. A strong signal fails the substitution: the description says something that is true of this company and false of others, something a buyer could check, repeat, and act on. I have run this test in brand workshops and watched the room go quiet, because the description that failed was the same description the company’s own homepage leads with. The emptiness was not the model’s invention. The model had learned the brand from the brand.

I have seen answers that mentioned a company without describing it positively, and answers assembled from odd pages. Observing that is not proof that one missing anecdote or one added sentence caused it. The practical response is at the source: strengthen the foundations, publish original material from real experience, earn credible external mentions, and keep the public descriptions consistent. The answer layer can only work with what the public world supplies.

[AWAITING MIKLOS — scene: second scene]

Miklos Roth

What question tests distinctiveness inside the answer?

Brand measurement asks whether the company is mentioned and how favorably — quantity and sentiment. The question that tests signal asks what the buyer can do with the description: can they repeat, to a colleague, one specific thing this company is for? A brand that passes that test has signal; a brand that fails it has presence with good sentiment, which is the condition I keep finding. The test is brutal, cheap, and almost never run, because brand dashboards were built before the answer layer made it matter.

The question that measures the signal: has this company supplied the public world with a clear reason to remember it — and could a stranger, reading what is publicly available, repeat that reason accurately?

PR excerpt (80–120 words, Miklos’s own words): What I count as a brand signal: familiarity, a consistent footprint of external discussion, and content distinctive enough that a reader can say what the company does and for whom. A weak signal is not a small one — it is an unclear one. If the public record gives the answer layer nothing specific to repeat, the company is not being rejected; it is being forgotten at scale. The practical response is at the source: strengthen the foundations, publish from real experience, earn credible external mentions, and keep public descriptions consistent.

© Copyright legolcsobb