From a single-page order form to Southeast Asia’s pharma commerce platform

eZRx is how hospitals, clinics, pharmacies and wholesalers across the region order medicines from Zuellig Pharma. In 2021 it did one thing: capture an order.

The product had grown around basic transaction fulfilment. Customers struggled with login friction, thin product information and an order status they couldn’t trust, so many simply called a sales rep instead. Digital adoption sat at 24%, and the business wanted a self-service channel that could work in every market.

A −36 NPS told us customers were unhappy. It didn’t tell us why. So before touching a screen, we went looking for the cause.

We took the 502 open-ended comments behind the NPS score and affinity-mapped them into themes. In parallel, a 462-person survey asked one focused question: what information do you need most when you place an order?

Two findings shaped everything that followed. Speed and downtime dominated the complaints, so this was partly an engineering problem, not only a design one. And the information customers wanted (expiry date, SKU-level name, pack size) was clinical and logistical, not marketing. Pharma procurement behaves differently from a generic shopping flow.

Talking to market teams surfaced four structural forces on top of that:

Rather than commit eleven markets to a redesign, we built a B2B commerce proof of concept to test one question: can a modern e-commerce experience support how pharmaceutical buyers actually purchase?

The work was organised around behaviour, role and job-to-be-done, not around screens. That meant aligning developers, market stakeholders and SAP use cases before a single pixel was finalised.

I built service blueprints with a bird’s-eye view for regional alignment and drill-down versions for micro-level decisions. They became the shared map that let market teams, developers and SAP specialists argue about the same thing.

I recruited six UI/UX designers from Deloitte for a three-month sprint focused on the three flows that mattered most: order, payment and return. Every week had a job.

The scores were good. The quotes were more useful, because they told us what “good” still left out.

That last quote stuck with me. Adoption doesn’t depend only on the person placing the order; it depends on everyone in the customer’s organisation who touches it.

eZRx Plus launched in Malaysia and Singapore as pilots. As it expanded, the in-house team I led took over from the agency and the work shifted from “does this work?” to “does this work here?”

Payments and returns had never been online, and every market ran them differently. Before designing them, I mapped the real service operations end to end, down to the bank responses and warehouse handovers.

Remote testing had limits. We went into pharmacies, clinics and warehouses across nine markets and watched people order where they actually work.

Some of the most important findings had nothing to do with the screen. Being on site showed us problems no prototype could.

User testing raised hypotheses. I used Mixpanel to check them against real behaviour across all markets before we spent space on the screen.

Scaling to eleven markets meant the details had to hold everywhere: multi-language UI, a three-step maximum for navigation, WCAG 2.0 AA contrast, consistent UX writing, and a chatbot conversation design so customers could ask for order totals or invoice status without calling anyone.

A pharmacy or hospital doesn’t just order from Zuellig. It pays invoices, sends stock back, earns rewards and chases the best deal. These modules bring that whole commercial relationship into eZRx, so customers stop picking up the phone for it.

eZRx wasn’t alone. Zuellig’s other products each carried their own library, and every one of them drifted. Consistency across eleven markets was impossible while the foundations were split 23 ways.

Remember the finance team that wouldn’t log in? These POCs take the product to where work already happens: inboxes, delivery routes and credit desks.