For twenty-five years the physical shop has been described as a cost centre in managed decline — a legacy channel to be rationalised, converted into a fulfilment node, or given an experiential veneer to justify its rent.
The argument was sound while its premise held. If a store's function is to hold inventory near customers, then a warehouse and a courier do it better. But that was never the whole of what a shop did, and the part that was overlooked has become the valuable one.
The evidence problem
Consider what a customer is actually doing when they evaluate an unfamiliar seller online. They cannot inspect the product, so they look for proxies: photographs, descriptions, specifications, reviews, the general impression of professionalism. Every one of those proxies is now cheap to fabricate convincingly and at scale.
This is not a prediction. Photographs of products that do not exist, reviews written by systems that never bought anything, and specification sheets generated to match a search query are ordinary features of the current market. The customer's instruments for telling a real seller from a plausible one have been degraded roughly to the point of uselessness.
A shop cannot be generated. It occupies a specific address, it pays rent to a landlord who exists, it employs people who can be spoken to, and it can be visited by anyone who doubts it. These are not romantic properties. They are epistemic ones: a physical location is expensive, and expense is what makes a signal hard to counterfeit.
What this changes about store economics
If the shop is partly an instrument of proof, then measuring it as a sales channel understates it in a specific and correctable way.
The store's contribution includes the online orders placed by people who were reassured by its existence without ever entering it — a group that conventional attribution cannot see at all, since visiting a website after knowing a brand has premises leaves no trace distinguishable from any other visit. Firms that have run market-level tests, opening or closing physical presence and observing online demand in the surrounding area, consistently find effects that store-level profit and loss statements do not capture.
This does not rescue every unprofitable store. It does mean that the closure decision, made on a store's own numbers, is being made on the wrong numbers.
Presence without retail
The implication is not that every firm should open shops. It is that the underlying function — being verifiably somewhere — can be served in more than one way, and firms that never considered themselves retailers should consider whether they need it.
A workshop that can be visited. A named address that is not a mailbox. Staff who appear under their own names rather than as a support queue. A factory tour, an open day, a showroom that sells nothing. Each of these does the same epistemic work as a shop: it converts a claim into something that could, in principle, be checked.
The test is simple. If a sceptical customer wanted to satisfy themselves that your business is what it says it is, what could they do? For a great many firms the honest answer is: nothing, except read more of the same material that made them sceptical.
The uncomfortable corollary
There is a version of this argument that firms will find attractive and should resist: the idea that the appearance of physicality will do. A photograph of an office, a map pin, a stock image of a team — these are the digital assets whose credibility has just collapsed, and adding more of them does not help.
What does the work is the possibility of verification, not its imagery. A shop that no one visits still functions as proof because it could be visited. A photograph of a shop functions as proof of nothing at all, which is precisely the change that has occurred.