Trade ordering that feels effortless.

A consumer store assumes one price, one basket and a card at the end. Trade assumes none of those things. Your buyer has a negotiated rate that differs from the buyer before them, an agreed credit limit, a purchase order number their finance team requires, and a catalogue restricted to what their account is approved for. On a platform designed for retail, each of those is a workaround - an app, a plugin, a custom field, a manual step - and the workarounds are what eventually make the store unmaintainable.

This is why trade stores so often fail quietly rather than loudly. Nothing breaks; the site simply never becomes the easy option, so buyers keep phoning their rep, and the store ends up as an expensive catalogue that reports 4% of revenue. The test of a B2B store is not whether it can take an order. It is whether your existing customers stop ringing.

We build on Medusa for exactly this reason: pricing rules, approval flows and catalogue permissions are things you build rather than things you request. Where a hosted platform genuinely fits your model, we will say so and save you the money - that honesty is on the e-commerce service page too.

Most B2B store traffic is not discovery. It is a purchasing manager buying roughly what they bought last month, under time pressure, who wants the whole task finished in under two minutes. That single behaviour should drive the entire interface: order history that can be reordered in one action, saved and shared lists for recurring requirements, bulk entry by product code for people who know exactly what they want, and stock visibility honest enough to trust.

The uncomfortable implication is that most of the design effort belongs to logged-in screens nobody sees in a pitch. The homepage matters far less than the account area, which is the reverse of a consumer build and the reason a consumer-shaped agency struggles with trade work.

Get it right and the return compounds in a way marketing rarely does: every reorder that happens on the site instead of on the phone is sales time returned to selling, and the effect grows with the customer base rather than decaying with the campaign.

Stock, pricing, accounts and orders already exist in your ERP. A trade store that is not connected to them is a second set of numbers to keep wrong, and it fails at the exact moment it matters: a buyer orders something that is not there, at a price that is not theirs. Connecting the store to the systems you already run is where most of the value and most of the risk live.

It is also where honesty about scope matters. Integration is priced after technical discovery against your real systems, never from a feature list - roughly £2,000 to £6,000 for a one-way connection and £6,000 to £20,000 for two-way synchronisation with business logic. The variable is rarely the API. It is data quality, edge cases, and decisions your business has not yet made about which system wins when both have changed the same record.

That discovery is worth buying on its own. A document that says "this is a £9,000 project, here is why, and here are the two decisions your team must make first" is useful even if you take it elsewhere.