Writing a post about a project when it has just finished forces you to take stock. To look over the last few months’ work and to reflect on what did and what didn’t work. We have just finished – last Friday – developing a knowledge sharing application for a large Professional Services firm for use by all of its UK employees.
The app’s purpose is to help their employees to understand the entire breadth of services offered by the firm. And to know the firm’s view on hot client topics so that they know enough to hold a conversation with a client on the topic and then know who to call about this opportunity. In language of ‘IT sales’, it is a pre-sales tool. Based on their predominant crop of devices, we built a Windows application (for use on Windows7 desktop) and an iOS application. Both synchronised with a purpose-built content and contact management system.
The applications needed to be intuitive, fast, work offline and offer a search experience that was tag, content and relevance based.
Lessons learned? We should have followed our business systems’ development methodology (more Agile) rather than our consumer one (Waterfall). That would have meant us starting development much sooner in the process – based on lo-fi wireframes and not hi-fi wireframes and signed off designs. Starting development sooner would have allowed more time for it, more time for data integration and more time for testing and iteration. Bringing forward the delivery date.
We would then be developing the system whilst conversations were happening about the design of the homepage….But as this project was initiated by the Marketing department, despite it being an internal application, we thought it made sense at the time to follow a waterfall approach to allow for sign off and stakeholder buy-in (both harder to do when being more Agile)
However at the end the project, you wish it had been treated more like a business system. Would it have looked much different and been better from a UX perspective? Possibly, but we could have been on Phase 1b by now with a set of enhancements based on the real user actions. Perhaps what I am saying is that sometimes it is worth the client sacrificing the comfort of a UX and Design driven, sequential step process for earlier development, greater testing and an earlier release date. But – and it is an important ‘but’ – only as long as there is guaranteed scope for post-live user oriented enhancements.
It takes more a little more bravery on a client’s part to take this approach, which is ironic really as clients are seemingly much happier to take it when developing a system that actually runs their business than one that portrays it.