malvasia bianca

A week ago, we finally finished all the cards that we’d planned for the week. I am hugely embarrassed that it’s taken so long, given that we’ve been trying to do the planning game for about ten months. But we’d been getting it pretty seriously wrong for a long time – reading through that post, all I can say is that I was really ignorant and must have been wearing some seriously rose-colored glasses. For the last couple months or so, though, we’ve been doing it better.

Not enough better, though, one might claim, given that we never managed to finish all the cards planned for a week. We frequently finished almost all of them as well as some unplanned cards, so our velocity hadn’t been going down, but there was always one card that we couldn’t do for external reasons. There was one partner integration test that took forever to do while cleverly stringing us along, thinking that we were always one answer from our partner away from getting it working. So we kept planning that card every week, thinking that this would be the week we’d finish it. And we kept on not finishing it, for week after week; I think it kept us going for two months solid. And then once we had that done, there was another card with external dependencies that strung as along for a few weeks as well. Eventually, we realized that we just had to wait for the external dependencies to be satisfied, so we left that card off the schedule for a few weeks, giving us our opening.

When I noted in our weekly meeting that this was the first week where we’d finished all the planned cards, one of my team members said “it’s because all the cards were 1-point and 1/2-point”, and I think there’s a lot of truth to that as well. We can deal with 2-point cards, but they give us less margin of error; smaller cards are much easier. (Unfortunately, we have a lot of 2-point cards for the rest of this month; I’ll try to avoid that in the future.) One way in which we aren’t doing strict XP is that not all of our cards make independent business sense: we’re not always clever enough to be able to come up with small cards that do make independent business sense, and given the tradeoff between small cards and independent business sense, we’re chosing the former. (Quite correctly, I think.) Perhaps playing into this is the fact that the person playing the Customer role is more technical than ideal, so we have less pressure on that front than some teams would. Though, actually when we do the monthly release planning (he’s not involved in the weekly planning unless some sort of surprise or prioritization judgment call comes up), he frequently doesn’t look at the individual cards: he plans things more a feature at a time. Which is nice: our monthly releases are pretty coherent, which is where it matters. And they’re starting to get quite predictable, which is great.