Your reputation is written where you can’t edit it

Most of what search quality looks for, you can put on your own page. You can write expert content, state your author’s credentials, lay out your purpose on a clear About page, and structure the whole thing so a reader and an engine can see what you are and what you know. Expertise and authoritativeness can, to a real degree, be shown on the site you control — demonstrated in the quality of the work, the depth of the content, the visible credentials of the people behind it. They are not entirely self-asserted either, but a well-built site can make a genuine case for them. Trust is different in kind, not just degree: there is no paragraph you can write that establishes it, because the thing being measured is precisely the part of your standing that exists in other people’s judgement rather than your own presentation. Trust is the part that resists this, because the heart of trust is reputation, and reputation is by definition the view others hold of you — not the view you hold of yourself. Google’s quality rater guidelines draw exactly this line: when they assess reputation, they direct raters away from the site’s own claims and toward what independent sources say. It is the one quality signal you are structurally barred from writing, and that is not an oversight in the guidelines; it is the whole reason the signal means anything. A measure of trust that you could set yourself would measure nothing but your willingness to claim it, and everyone is willing to claim it. By placing reputation outside your reach, the guidelines make it the one part of the quality picture that cannot simply be asserted into being, which is exactly why it carries the weight it does in the assessment.

The value of reputation as a signal comes entirely from the fact that you did not produce it. A glowing self-description proves nothing, because anyone can write one; a pattern of genuine assessments from people and publications with no stake in flattering you proves a great deal, because they had no reason to say it unless it was true. This is why the guidelines insist on independent sources — third-party reviews, references in articles, recognition from others in the field — and why a testimonial you selected and published on your own site does not count as reputation evidence in the same way. The independence is the load-bearing word. A review on a platform you do not own, written by a customer you cannot edit, sits in a different evidential category from the same praise quoted on your homepage, because in the first case you could not have stopped an unflattering version from appearing and in the second you could. The rater is trained to notice which kind they are looking at, and to discount the kind you curated. The independence is not a bureaucratic preference; it is the entire mechanism by which reputation carries information. Strip the independence away and you are left with marketing, which the rater has learned to read as marketing and to weigh accordingly. The same sentence carries entirely different evidential force depending on who wrote it: a description of your expertise is persuasion when you write it and evidence when a respected outside source writes it, even if the words are nearly identical. That is not unfair; it is the only way a reputation signal could possibly work, since the entire informational value lies in the absence of your hand on the pen.

The sharpest expression of this principle is what the guidelines say to do when the two sources conflict. If what a site advertises about itself contradicts what independent sources report, the rater is told to believe the independent sources. Your own account does not get the benefit of the doubt against a credible external record; it loses to it. That is a humbling rule if you have spent effort polishing the story you tell about yourself, and a clarifying one. It tells you, in effect, that your marketing department does not get a vote that outranks your customers. The asymmetry is deliberate and total: the external account is treated as the more reliable witness on the question of whether you are trustworthy, precisely because it has less reason to flatter. And it is clarifying because it tells you exactly where to spend the effort — not on the telling, which cannot win the argument, but on the being, which is the only thing the witnesses are reporting on. Once you internalise that, a great deal of anxious effort can be redirected. The hours that might have gone into perfecting the language of your own trustworthiness are better spent making the underlying thing more true, because the language was never going to be read as evidence anyway. There is an odd relief in it: you are released from the impossible task of arguing yourself trustworthy and handed the merely difficult one of being so, a task you can actually complete through your own conduct. it means the self-description can never out-argue the reputation, only at best agree with it. The practical consequence is that the gap between how you present yourself and how you are actually regarded is not a gap you can close with better copy. You close it by changing the thing the outside sources are reacting to, which is your conduct, not your claims. This is a slower and less comfortable lever than rewriting a page, because it means the only reliable way to improve how you are spoken of is to become more worth speaking well of, and then to wait while the external record registers the change. There is no shortcut through the copy, because the copy is not where the verdict is rendered. The verdict is rendered by people you served, in places you do not own.

It would be easy to read all this as a demand that every site arrive with a thick file of glowing external coverage, and the guidelines are careful to say it is not. For small businesses and organisations, an absence of reputation information does not by itself make a site low quality. A new operation, a niche practice, a business in a field where customers simply do not leave reviews — none of these is penalised merely for the quiet. The guidelines treat the absence of reputation information as a neutral fact about a small or young entity rather than as a mark against it, which is both fair and practically important, because otherwise every new and honest operation would start with a quality deficit it had done nothing to deserve. The bar is not a thick file of praise; the bar is the absence of a credible bad record. That is a far more reachable standard than it first appears, and it should reassure the small and the careful rather than alarm them. You do not need to have been written about widely; you need to not have earned a credible reputation for harm, dishonesty, or incompetence. Most honest operations clear that bar simply by being honest, and the reputation then grows from there at whatever pace their work and their relationships allow. Quietly competent and not yet much discussed is a perfectly respectable place to be, and the guidelines say as much. The distinction that matters is between having no reputation yet and having a bad one: silence is survivable and expected for the small and the new, while a credible negative record is the thing that genuinely counts against you. What you must not do is try to fabricate the missing signal, because invented reputation is precisely what the independent-source test exists to catch, and getting caught manufacturing it is worse than the silence you were trying to fill. Fabricated reviews, staged testimonials dressed as independent, and reputation laundered through sites you quietly control are all failures of exactly the test the guidelines are applying, and the penalty for being seen to fake the signal is heavier than the mild, survivable cost of simply not having much of one yet. A thin record invites patience; a faked one invites the conclusion that you are not to be trusted, which is the precise opposite of what you were trying to buy.