When most people think about product development, they think about metrics: DAU, conversion rate, churn, ARR. The incentives of commercial software have shaped how we design, what we measure, and how we define success.
Building UMAS — a research platform for the Faculty of Pharmacy at Universitas Padjadjaran — forced us to think differently. Academic platforms have a different success model, different users, and different constraints. Here's what we learned.
Success Looks Different
In commercial products, you know immediately when something works. Users sign up, return, pay, refer. The feedback loop is tight.
In research platforms, success is measured in citations, in studies enabled, in researchers reached. A researcher in Ghana downloads the instrument and uses it for a study that gets published two years later. You won't know that happened unless they tell you — and they probably won't.
This changes how you design. Instead of optimising for short-term engagement, you optimise for trust and accessibility. Will a researcher trust this platform enough to use its instrument in a peer-reviewed study? Will they be able to access it reliably from a slow connection in a resource-limited setting?
Credibility Is the Product
For academic users, the platform itself is a signal. A poorly designed website casts doubt on the rigour of the research it hosts. A clean, professional, well-structured platform communicates that the work behind it is equally serious.
This is different from building for consumers, where personality and warmth matter most. Academic users want clarity, precision, and evidence. Every design decision has to earn its place by serving those values.
Open Access Is a Design Decision
The UNPAD team made a deliberate choice: UMAS would be open access. No paywalls, no institutional subscriptions, no waitlists. A researcher in a low-income country should have exactly the same access as one at Oxford.
That decision had design implications. The request flow needed to be simple enough that access wasn't a barrier in itself. The platform needed to work on low-bandwidth connections. The language needed to be accessible to non-native English speakers.
Open access is not just a business model. It's a value — and when you share it, it shapes every detail of what you build.
Longevity Over Novelty
Research platforms need to last. An instrument validated in 2024 might be cited in studies published in 2034. The platform needs to still work then — still be findable, still serve the same purpose, still reflect the credibility of the work.
This pushed us toward conservative, stable technology choices over anything cutting-edge. Next.js, well-structured HTML, clean URLs, solid SEO fundamentals. Nothing that would require a rewrite in two years.
Building for academia taught us that sometimes the most sophisticated choice is the most boring one.