The storefront was never the moat

Fifteen years ago a good online shop was a competitive advantage. Building one was expensive, running one was hard, and plenty of retailers had a bad one or none at all.

That is over. A competent storefront is now a subscription and a fortnight, everybody has one, and yet it still takes the largest share of retail software budgets — a redesign every three years, a replatform every six, and a great deal of attention paid to something a competitor can match in a month.

What actually differentiates

Knowing what you have. A retailer who can promise accurately sells more of what they own and disappoints fewer people. This is dull, hard, and almost never a project.

Knowing your product. Structured knowledge — what fits what, what a thing is for, what it will not do. This is what a good shop assistant has and almost no catalogue does, and it is very hard to copy because it accumulates.

Knowing your customer, honestly. Including returns, complaints and the things they bought elsewhere afterwards.

Getting goods to people. The physical part, which is genuinely hard and genuinely a moat, and which no storefront redesign affects.

Buying well. The oldest advantage in retail, and unchanged by any of this.

None of these are visible on a homepage, which is precisely why they last.

Why the budget goes to the wrong place

It is visible. Everybody in the business has an opinion about the shop and nobody has one about stock reservation semantics.

It is projectable. A redesign has a start, an end and a launch. Improving stock accuracy is a programme with no announcement at the end.

Vendors sell it. There is an industry selling storefronts and almost nobody selling be more accurate about your stock, because the second is not a product.

It feels like the customer-facing part. It is the customer-facing part. It is also the part where everybody is roughly equal.

The uncomfortable test

Ask what a competitor would have to do to match you. If the answer is buy the same platform and hire the same agency, that is not a moat — it is a purchase order.

If the answer involves five years of product knowledge, a warehouse process that works, and a stock figure people trust enough to promise from, that is harder to copy, and none of it lives in the shop layer.

What that implies for where a system sits

This is why Mercantir is underneath the storefront rather than replacing it. Most retailers have a shop that works and do not need another one. What they usually lack is a truthful record of stock, orders and product knowledge for the shop to draw on — and for whatever the next channel turns out to be.

Channels change. Marketplaces appeared, then social selling, and something agent-shaped may be next. Each one arrives assuming you can answer what do you have, what is it, and when can it be here. A retailer who can answer those is ready for a channel that has not been invented yet, and one who cannot will rebuild the same storefront again.