Over the years, the design of MediaWiki has been solely driven by software developers. This has caused an unfortunate technology-based approach of the front-end and the features (implemented or missing), relying mostly on the implementation model. The consequence is that the interface & features are too far from the users’ mental model.
Japanese Tea pot by Denis Savard, under CC-By, from Wikimedia Commons.
UX should be a systematic approach
The Wikipedia and Multimedia Usability projects have tried to address the most pressing concerns resulting from the hiatus between the software and the users’ expectations. For all these reasons, I am really happy to see the Wikimedia Foundation investing further in User Experience (UX). However, I see little added value in having a UX department separate from the main development cycle.
A more systematic approach is necessary in order to improve the usability of Wikimedia projects perennially; good, usable design needs to happen before the actual implementation of any feature, in the early stages of the product (or component) development. Otherwise, we will always be running after the train, and never catch it.
A separate group made sense when these UX programs had a specific scope and time frame, but it was because they were tied to specific grants. In a more permanent setup, I see no reason to separate UX programs from the “regular” development processes; targeted actions can be carried out by specific projects inside the development team, rather than by a separate team altogether.
Everything is UX
More generally, all the activities of our Technology department are about User experience; everything we do is UX. Software development aims to fix bugs, develop new features, improve others, and remove hindrances. The sole goal of all of these activities is to improve the user experience by making the software better and closer to users’ needs. Even Operations are about UX: the goal of the Operations team is to make sure the information can be accessed reliably and reasonably fast by an audience as large as possible; in short, the point of Operations is to ensure we actually provide a user experience.
As a consequence, I recommend to make UX a systematic part of the product or component development cycle, not a separate parallel group.