Castor Coffee sells both to consumers and to businesses. Those two buyers differ in price, basket, tone and even which products they should see. Instead of forcing one store to do both, two connected stores were built on WordPress multisite.
Two separate WooCommerce stores on a single WordPress multisite installation: a public retail store and a B2B/wholesale store visible only to logged-in businesses. The retail side gained a product personalisation (label) configurator and a blend-building tool that recommends by intended use rather than by bean. An in-café POS system and corporate email infrastructure were also set up for the same brand.
Coffee roasters usually run two businesses at once: selling bagged coffee to consumers and supplying cafés, restaurants and offices wholesale. The two share products and almost nothing else.
Squeezing both into one store classically breaks both: layer upon layer is added to hide prices, cart rules grow complicated, and in the end neither buyer is served.
The build sits on WordPress multisite: one installation, one plugin and theme pool, one admin — but two independent stores.
The B2B store is closed to visitors who are not signed in. Wholesale prices are not visible until a business account is verified.
The commercial logic is simple: a wholesale price is a quote, not a public label. The technical logic concerns search engines — a gated B2B store is not expected to appear in Google, and should not. Brand visibility is carried by the retail store and the corporate pages; the B2B side is a tool, not a marketing surface.
The retail side gained a configurator that lets a customer personalise the product's label or sticker — a clear differentiator for gifts and corporate orders.
The easiest mistake in a configurator like this is turning it into a design application. The more options you offer, the more decisions the user must make and the more often they abandon. So the tool was deliberately kept narrow: one product, full width, a small number of meaningful choices. The goal is not infinite freedom but a good-looking result in a few clicks.
Most people buying coffee do not know which blend they want. They cannot translate "make it strong", "good with milk" or "low acidity" into a product name. That is the most common stall point in coffee e-commerce.
So a build-your-own-blend tool was developed. Rather than presenting a list of beans, it works from intended use and taste preference and recommends a blend. The question is not "which bean would you like" but "how will you drink this coffee".
It is a design decision that takes the decision load back from the customer: expertise should sit with the brand, and the interface should reflect that.
A POS system was also built for the café floor, so in-store and online sales are managed as two arms of one brand.
POS is a separate system because a café till works under completely different conditions from e-commerce: it must be fast, touch-driven, error-resistant and tolerant of poor connectivity. Loading both workloads onto one interface weakens both.
Castor Coffee now runs retail and wholesale on the same WordPress installation as two stores that do not undermine each other. Personalisation and the blend tool serve the brand; POS and corporate email serve operations.
The transferable lesson: do not squeeze B2B and B2C into one store. The catalogue can be shared; price, access and interface cannot. Multisite is the most practical way to build that separation without doubling maintenance.
Technically yes, but it rarely works well. You end up hiding wholesale prices, applying different minimum-order rules and designing one interface for two very different buyers. Two stores on WordPress multisite remove that conflict while keeping a single maintenance point.
If both stores belong to the same brand and share product logic, multisite is the better fit: updates happen once, the security surface stays single, and users and roles are shared. For two genuinely independent businesses, separate installations are cleaner.
No — that is intentional. Wholesale pricing and dealer terms should not surface in search results. Search visibility is carried by the retail store and the corporate pages; the B2B side is an ordering tool for existing partners.
Noticeably so for gifts and corporate orders, because it makes the product personal. But the effect reverses as options multiply — the user cannot decide and abandons. Build the configurator with the smallest number of meaningful choices possible.