What brand mention building is
Brand mention building is the deliberate pursuit of being named — in articles, forum threads, review roundups, podcasts, industry reports, newsletters, community answers and reference pages — whether or not a link comes with the mention. It is the same work as digital PR and expert commentary, pointed at a different target. Instead of asking what will earn a link, you ask what will get your name into the corpus of text that describes your industry.
This sounds like a rationalization for failing to get links. For most of the last twenty years it would have been. What changed is how search systems read text. A search engine no longer needs an anchor tag to know that a phrase refers to a company; it can resolve the string to an entity, associate it with the topic of the surrounding sentence, and count how consistently other sources make the same association. Large language model based answer engines do something similar and more directly: they are trained on and retrieve from text, and text is where your name either appears or does not.
The unlinked mention has therefore moved from being a lead — something to chase and convert into a link — to being an outcome with value of its own. Both readings are still true. That tension is the whole subject of this page.
How it works, step by step
- Define the entity you want established. One brand name, spelled one way, attached to two or three specific topics. Scattered naming — a legal name, a trading name, three product names — dilutes everything that follows.
- Find the sources that describe your category. Trade publications, comparison and roundup articles, subreddits and forums where practitioners answer each other, industry newsletters, association reports, and the pages that already rank for the questions your buyers ask. These are the texts that get read, cited and retrieved.
- Give those sources a reason to name you. That is almost always one of four things: original data nobody else has, a named human with a defensible opinion, a genuinely distinctive product or method, or availability at the moment a journalist needs a quote.
- Be present where practitioners talk. Not astroturfing. Answering questions under your own name in communities that allow it, and accepting that most of those answers will never carry a link.
- Make your own site easy to reconcile. Consistent naming, a real about page, an author page for every expert, clear descriptions of what you do in plain sentences. A system that cannot confirm who you are will not repeat the claim.
- Reclaim the ones that should have been links. A polite note to an editor converts a share of mentions into links. Do it — but treat the conversion as upside, not as the point.
What it costs in time and effort
The cost profile is the awkward part. Mention building consumes the same expensive inputs as digital PR — research, a credible spokesperson, journalist relationships, responsiveness measured in hours — but returns a deliverable that is harder to put in a report. You are paying senior-level time for an outcome that a spreadsheet cannot easily total.
There is a second, subtler cost: patience. Entity associations build slowly and through repetition. One mention in a trade publication changes nothing. Forty mentions across three years, all attaching the same name to the same subject, change quite a lot. Nothing about that timeline fits a quarterly reporting cycle.
What it does not cost is risk. There is no version of this work that produces something to disavow. That matters more than it sounds: a large part of the link building budget in this industry is spent on tactics whose downside is a cleanup project two years later.
When it works and when it does not
It works when there is something real to name. A company with proprietary data, a distinctive method, a recognizable expert or an unusual product gets mentioned because writers need concrete examples. It works especially well for professional services, B2B software, specialist manufacturers and anyone whose buyers research by asking other people rather than by searching a product name.
It does not work for undifferentiated resellers. If your business is indistinguishable from four competitors, no amount of outreach will make a journalist name you specifically, and no amount of naming will make an answer engine prefer you. The mention has to be earned by the thing being mentionable.
And here is the honest part. If your reporting line counts links, this tactic will look like underperformance for a long time before it looks like anything else. I would not recommend selling it as a replacement for link acquisition to a client who has never bought the argument. Sell it as what it is: a second, slower asset class that reduces your dependence on a link graph which is itself decaying under you. In my own profile, 49.6% of every link ever recorded is already gone, and the median referring domain stops linking after 1,080 days. Anything you build that does not evaporate on that schedule is worth having.
Common mistakes
- Using mention building as an excuse. If your campaign was supposed to earn links and earned none, that is a failed campaign, not a mention strategy. Be honest about which one you ran.
- Chasing volume of the wrong mentions. Ten thousand automated social posts naming your brand are not citations. The mentions that matter appear in text that other people read and that search systems retrieve.
- Inconsistent naming. Being called four different things across a hundred sources builds four weak associations instead of one strong one.
- Astroturfing communities. Fake accounts recommending your product are detectable by moderators, by readers and increasingly by the platforms themselves. When it is caught, the association you built is negative.
- Ignoring reclamation. The easiest links available to most sites are mentions that already exist without one. Not asking is leaving money on the table.
- Reporting it as if it were links. Mention counts presented in a link report invite exactly the comparison you will lose. Report them as their own line with their own rationale.
A worked example
Consider a specialist insurance broker in a technical niche. It has no budget for a large content operation and no products that generate natural links. What it does have is a partner who has settled several hundred claims of a specific kind and can explain, precisely, why they fail.
The program is narrow. That partner writes a short, genuinely useful explanation of the claims failure pattern on the firm's own site under her own name and photograph. She registers with journalist request services and answers only questions in that one area, within the hour, in usable prose. She takes podcast invitations. She answers questions in two professional forums where practitioners discuss these claims, under her real name, with no links.
Twelve months in, the link count is modest — some journalist request placements, a couple of trade features. The mention count is not. Her name and the firm's appear together across trade press, podcast show notes, forum threads and roundups, always attached to the same subject. When somebody asks an answer engine who understands that claim type, the corpus has a consistent answer available to it. The links that arrive later arrive because that association exists, not the other way round.
How to measure it
Measurement is the weakest part of this tactic and pretending otherwise helps nobody. What can be measured:
- Mention volume and source quality. Track named mentions across monitoring tools, separating publications and communities that matter from noise. Grade sources the same way you would grade a link prospect.
- Consistency of association. Are the mentions attaching your brand to the two or three topics you chose, or to nothing in particular?
- Share of voice against named competitors in the same set of sources, tracked over quarters rather than weeks.
- Presence in generated answers. Ask the same set of buying questions across answer engines on a fixed schedule and record whether you are named, and how you are described. It is crude, the results move around, and it is still the most direct signal available.
- Conversion to links. What share of mentions became links after a reclamation request. This is the number that keeps a skeptical stakeholder in the room.
Do not manufacture a single composite score out of these. A fabricated index is worse than an honest set of imperfect measures.
The verdict
Safe, slow, and increasingly the thing I would prioritize if I could only run one earned-media program. The reasoning is not that links stopped working. It is that the same activity now produces two assets instead of one, and the second asset — a consistent, widely repeated association between your name and your subject — does not decay on the schedule links do.
Be honest about the trade. This is the hardest tactic on the site to defend in a report, and anyone who tells a client that mentions are simply the new links is overselling a position the evidence does not yet support that strongly. Run it alongside link acquisition, reclaim the mentions that should have been links, and let the mention count be a measured secondary outcome rather than an alibi.
Questions
Do unlinked mentions really help rankings?
The honest answer is that nobody outside Google can quantify it. What is defensible is the mechanism: search systems resolve brand names to entities and use the text around them to understand what a company is known for, and answer engines retrieve from text rather than from anchor tags. Treat mentions as an established input of uncertain weight, not as a proven ranking factor.
Should I still ask for the link when I find a mention?
Yes. A mention that already exists is the warmest link prospect you will ever have, the request costs one short email, and a reasonable share of editors will oblige. The argument for mention building is not that links stopped mattering. It is that the mention retains value when the request is declined, which it often is.
How is this different from unlinked mention reclamation?
Reclamation is reactive: you find mentions that already happened and try to convert them. Mention building is proactive: you do the work that causes mentions to happen in the first place, and you count the ones that never carry a link as a result rather than a miss. They are the two halves of the same program.
How do I report this to a client who only counts links?
Do not disguise it. Report links as links, mentions as mentions, and show the conversion rate between them. Then make the argument on durability: links decay, and roughly half of every link my own site ever earned is already gone. A named association in widely read sources is an asset with a different decay curve, which is why it is worth funding separately.
Does social media chatter count as a mention?
Barely, and less than vendors selling mention tracking would like. A mention has value when it sits in text that people read, that gets indexed, and that other sources draw on. A post in a feed that scrolls past in a day rarely qualifies. Forum threads, newsletters, trade coverage and reference pages generally do.