Scrum: Day 1

Today saw the kick-off of two very different projects using Scrum. Having decided to try Scrum, I’ve adopted the philosophy and become a pig – it’s my bacon on the line – as I’ll be explaining to the MD tomorrow.

The first project is an ongoing development to build an interface with a third party. On Friday we had hit the first major milestone: just enough functionality that the third party can test our interface. We had a long list of improvements to the API and the UI to make the data which comes from the interface usable. This first sprint was time boxed as a week long, a quarter of the standard sprint time but a time scale we’re more used to. I assembled the project team: my lead developer, a junior developer, a tester, the product owner (who’d been introduced to his new role on Friday) and the necessary chicken: one of my peers. The kick off meeting started slowly until the team started to take over, prioritising requirements and designing solutions. One of things that suprised me was how some hidden requirements came out – the results of commercial discussions that hadn’t made it as far as a specification document. If Scrum achieves nothing else then it has already improved communications between the business and development as well as giving my developers some buy in to – and influence on – what we build. It’s easy to see how a cycle of kick-offs and reviews will build a mutual understanding of the requirements and their priority.

The second project is to modify a web site; to take some unused functionality out, tweak the design, add some new functionality and restructure it a bit while we’re there. I’d estimated the whole thing would take 4 man days of development, 2 days of design time and a day’s testing. The account manager responsible for the project spends a lot of time writing specifications, briefs and wireframes. Most of my internal clients would provide a half-written badly spelled email; this account manager, by contrast, sends 20 page specifications and Powerpoint presentations. As the kick off started it became rapidly obvious that the initial estimate was too short. We split the work into two sprints and set about prioritising the requirements. By the end of the meeting we’d decided what would go into the first sprint and how we set about doing it. We left the meeting with a really good understanding of what is about to be developed. Tomorrow will show how accurate that understanding is…

  1. No comments yet.

  1. No trackbacks yet.