Scrum at FP: Sprint 9
So, we've just completed our 9th fortnightly sprint at Future Platforms, it seems like as good a time as any to write an update on where we are.
We had an excellent session at the end of Sprint 8: the team (nearly) unanimously asked for standups to start earlier in the day. We had been running them at 09:45 - basically because I'd felt, despite not being a morning person myself, that this is a reasonable time to be in, thoughts collected and ready for work. So I was a bit surprised to hear everyone wanting to get cracking earlier - but no complaints :)
One other thing that came out of sprint 8 was that the review section of the day (where we run through work from the last 2 weeks) didn't seem too useful: we were asking each member of the team to show off some work, which led to a few folks demoing bug fixes (not that interesting for the rest of the team), and our in-house QA showing off things that look normal. So we resolved to change the format of this section of the day - see below.
Sprint 9 wasn't so good for us. We under-performed significantly compared to our work rate on previous sprints. There wasn't a clear reason why this might be, I suspect a combination of losing 2 members of the team who are consulting full-time for another client (and are therefore less available for project work), and the sprint being 2 days short thanks to a Bank Holiday falling in the middle. We'd factored in the raw loss of time the latter would cause, but not the loss of rhythm.
Joh's been down with the lurgi, so I took on the task of facilitating the sprint 9 retrospective/sprint 10 planning day. The review went much better - this time around, we had our Product Owner (currently Sergio) specifically request what he wanted to see demonstrated... and the whole affair felt much more coherent. QA presented an overview of all the projects we're working on (bug counts, progress, etc.) which seemed worthwhile too - we're going to keep both of these amendments for next time around.
What came up in the retrospective? As they go, this one was quite negative: the pair consulting for another client have felt disconnected from the rest of the company and didn't see much value in contributing to standups where their work has nothing to do with anyone else - particularly when, as it happens, they attend another (Skype-facilitated) standup with the clients team. So we've excused them from these events for the rest of their time consulting - and done the same for anyone else who's working solo and dedicated to a single project.
A few technical items for discussion came out in the retrospective, which we'll be gathering to discuss - mainly specific to MIDlet development or specific projects, so I won't go into them here.
The debilitating effect of interruptions was once more noted - we suggested that team members who repeatedly experience this attend standups and offer themselves as available after these standups - a bit like "surgery hours". We'll see how this goes.
We need to think through how we run projects with external contractors.
Lunch followed, then planning - slightly frenzied as we rushed to gather together the fortnights work. The planning session itself involved only the development team (as they're the only folks collaborating on anything in this sprint, all our design team are on solo projects) and was probably the quickest and least painful we've had to date. We completed planning in around 90 minutes, which is a record. As for why? We had fewer people in the room this time around, the projects we were planning out are reasonably well understood and we're clearly focusing on a couple of areas this time around.
So there we are... better reviews, better planning, but uncharacteristically poor recent performance and still questions over integrating design and development disciplines into a coherent unit.