A company I know well spent a year rebuilding its website and rewriting its positioning. The new site was better in every measurable way. Traffic held steady. Inbound enquiries fell. What had changed was not the site. Six months earlier the company had quietly retired its old customer forum, a slow, ugly, publicly indexed archive where buyers had argued about implementation problems for the better part of a decade. That archive was doing more selling than the homepage ever had, and nobody had it on a dashboard.
That is the shape of the problem I want to describe. Distribution used to mean access to attention: shelf space, ad inventory, a mailing list, a partner channel. You paid for reach and reach produced consideration. Consideration is no longer downstream of reach. It is downstream of evidence. Building trust with customers online is not a brand exercise that happens after the demand-generation exercise. It has quietly become the thing that decides whether the demand-generation exercise works at all.
Consideration is now gated, not won
Think about the last significant purchase you made on behalf of your company. You almost certainly did not evaluate every credible vendor. You assembled a shortlist first, and the shortlist came from somewhere other than the vendors' own marketing. A peer. A community. A search that returned a synthesised answer naming three companies. A procurement colleague who had heard a name before. By the time anyone read a vendor's website, the set was closed.
The important word there is closed. Reach used to buy you a seat in the consideration set. Now the set is assembled before your reach has a chance to operate, and the assembly is done on signals of reliability: who shows up in third-party accounts, who other people vouch for, who has a track record that is legible from the outside. If you are not present in that layer, buying more attention mostly means buying attention from people who have already decided.
This is why I think of customer trust signals as distribution rather than as brand. Brand is what people feel about you once they know you exist. Distribution is what determines whether they encounter you at the moment a decision is being made. Verifiable evidence of reliability now does the second job. I have written elsewhere about how the first thing a customer reads about you is no longer yours; the corollary is that the assets doing your distribution are increasingly assets you do not own and cannot edit.
What actually counts as a signal
Most of what companies file under social proof is not a signal at all. A signal has to be costly to fake. A logo wall is cheap — anyone can license a stock photo of a boardroom and put five logos under it. A testimonial written by the vendor and approved by the customer is only slightly more expensive. These things are not worthless, but they carry almost no information, and sophisticated buyers have learned to skip them the way they skip a pre-roll ad.
The signals that move a decision share a structure: someone other than you bore a cost to produce them, and the record of that is public and dated. A customer who writes a detailed review of a difficult implementation, including the parts that went badly. A public changelog going back years, with the awkward months visible. A support forum where your own staff answer questions in the open and sometimes say "that is a known problem and here is the workaround." An executive who has said the same thing about the business in public for five years, checkable against what actually happened. A specification, a warranty, a published security posture, a reference customer who will take a cold call.
What unites them is falsifiability. Each one could have been contradicted by reality and was not. That is the whole mechanism. Brand credibility is not a feeling you cultivate; it is an accumulated record of claims that survived contact with the world.
How companies destroy the signals they have
The destruction is almost never a decision to destroy anything. It is a series of reasonable tidying-up exercises, each defensible on its own terms, that together remove the evidence a buyer would have used.
The rebrand is the classic case. A company changes its name or its domain, redirects poorly or not at all, and severs itself from a decade of third-party references. Every article, every forum thread, every citation now points at a 404 or a generic homepage. The company still has the history; the market no longer has access to it. I have watched firms treat this as a marketing project when it is a distribution project, and price it accordingly.
Closing the forum is the second case. Public customer communities are irritating to run. They surface complaints, they need moderation, they generate content nobody controls. They are also, frequently, the single richest source of evidence a prospect can find that real people use the product and get help when it breaks. Shutting one down and replacing it with a gated support portal is a straightforward transfer of information away from the people deciding whether to consider you.
Then there is the archive purge. Old blog posts get deleted because they are "off-message." Case studies from a discontinued product line come down. The changelog starts at the current version. Each deletion is tidy and each one removes a dated, checkable claim that survived. A company with a visible ten-year record of shipping is in a different category from a company whose public history begins eighteen months ago, and the second company often used to be the first.
The subtlest version is the over-managed voice. A firm decides all public communication must run through communications review. Engineers stop answering questions. The executive who used to write candidly now publishes approved copy. Nothing false is said; nothing costly is said either. The output becomes indistinguishable from every competitor's output, which means it carries no information, which means it stops functioning as a signal. Safety and silence look identical from outside.
The asymmetry that makes this worth board attention
Trust signals accumulate slowly and disappear quickly. You cannot buy ten years of public consistency. You can lose it in a quarter with a migration, a purge and a policy. That asymmetry is exactly what makes something a balance-sheet item rather than a marketing line item, and I have argued that reputation belongs on the balance sheet for the same reason.
What follows practically is modest. Someone should own the inventory. Not the narrative — the inventory. Which third-party assets currently carry evidence about us? Where do they live, who controls them, what would break them? In diligence I have started asking a version of this, because the answer separates companies that understand their own demand from companies that have confused their marketing spend with their market position. It sits naturally alongside the other diligence questions whose answers predict trouble: not what the pipeline looks like, but why anyone is in it.
The second practical consequence is that any migration, rebrand or consolidation needs a signal-preservation plan with the same seriousness as a data-migration plan. Redirect everything. Keep the archive. Keep the dates visible. If you are retiring a community, export it and leave it readable. The cost is trivial. The alternative is deleting your own distribution and then wondering why the new site underperforms.
Generating new signals is mostly a willingness problem
Companies ask how to build more social proof. Usually the honest answer is that they are unwilling to bear the cost that makes a signal a signal. Publishing a real uptime history means publishing the bad months. Letting engineers answer publicly means occasionally being wrong in public. Naming what you do not do, and choosing which customers not to serve, means turning away revenue in a way that is visible. Every one of those is a cost, and the cost is the point. A claim that cost nothing to make proves nothing.
There is a governance version of this too. Boards that publish a clear sense of how they operate, that keep the same directors for meaningful stretches, that describe their business the same way in a downturn as in a good year — those are trust signals aimed at a different audience, but they work on the identical mechanism. Consistency over time, checkable from outside, expensive to fake.
Protect the boring things
If I had to reduce this to one instruction for an operating team, it would be: before you tidy anything up, ask who was using it as evidence. The forum nobody wants to moderate, the URL structure nobody likes, the decade of posts that no longer match the positioning, the support engineer who answers in public without clearance — these are unglamorous and they are frequently the reason you get considered at all.
We are all going to keep spending on reach, and reach still matters. But reach delivered to a buyer who has never seen independent evidence that you are reliable is expensive noise. The companies I expect to do well in the next few years are not the ones with the best-argued positioning. They are the ones whose claims have been publicly checkable for long enough that nobody needs to take their word for it.