Some Agile Updates

It's been over a month since I last wrote specifically about Agile hereabouts, so I guess an update is overdue.

My second stint as a Feature Lead has gone well so far, despite the new team's obvious apprehensive about my arrival. They were concerned for a couple of reasons. First, their previous Feature Lead was well-liked and had done a good job in a tough situation. He was very committed to the team, and thus they were reluctant to see him exit and some bozo from left field come in and take his place. Also, I'd had a couple previous run-ins with this team, while I was Feature Lead elsewhere, and so I suspect they thought that I was coming in to clean house and/or re-make the team in my own image, or something like that. Thus, things got off to a bit of a rocky start. I spent a lot of my initial days there simply building up trust, and trying to convince them that if they gave it an honest effort, I'd do likewise to help them not only succeed, but be perceived externally as doing so.

And the first half-month Iteration we did together still wasn't successful. Many of the team members struggled with doing task breakdowns and estimating, and their Product Owner had gotten very used to asking them to take on whatever adhoc work came his way, all of which added up to an under-delivery in the Iteration. But even having said that, they showed a lot of promise in those two+ weeks, and I was hopeful for better results to come.

Sure enough, in the second Iteration we corrected a bunch of stuff from the previous one, including making the Product Owner much more aware of the impact of any unplanned work he asked for. And lo and behold, that Iteration was not only 100% successful, but the team was able to pull in a couple extra legacy bug fixes that came in during the Iteration! I noticed an appreciable uptick in the teams' morale after that result, but also concern that it was a one-time anomaly that wouldn't be repeatable.

Last week we completed the third Iteration, and once again saw 100% success. This time there was a bit of luck involved, as the team skated very close to missing some of the deliverables but pulled it off thanks to very little of the accounted-for adhoc work actually coming to fruition. But on the other hand, there didn't seem to be as many extra hours needing to be worked as they'd been doing previously, meaning that perhaps the team is close to achieving a sustainable pace that can still lead to successful Iterations.

One of the challenges that we haven't tackled yet, but will have to soon, is the ability to chunk the work into smaller pieces that can be completed throughout the Iteration. This team still tends to think in terms of each Iteration being a short little Waterfall, with the testing being heavier in the latter days. I've consciously decided to let that ride for the first few Iterations, while they work on the basics of determining what tasks are needed, and approximately how long each will take. They're almost there, so maybe in the next Iteration I can work on taking a feature and having them conceive sub-features for it.

Outside of my work with that Feature Team, I also crafted - with help - a table showing some Characteristics of an Agile Team as they move through the forming, storming, norming and performing stages. That document was published to everyone in the company late on Friday, so I expect to hear some fallout from it this week, if anyone actually reads it.

We also kicked off a Retrospectives Working Group, to bring "Retrospective enthusiasts - or wanna-be enthusiasts" together to think of ways to improve our Retro processes. We've been in a rut lately where many teams are starting to abandon the Retro process because they think they don't need it anymore, or they're tired of the activities involved. This group needs to figure out ways to, among other things, get the message out that Retrospectives, in whatever format the team chooses to do them in, are something the team does for itself, not something they should consider as mandated from outside. The Characteristics document mentioned earlier reinforces this need for inspection and adaption, as the document's intended to be a yardstick that Feature Teams can use to measure their progress in becoming Agile.

And lastly, I've been working with PeterJ to get "one-click e-mail subscriptions" to our work blogs and forums rolled out, so that more than the handful of readers we currently have can follow everything that's discussed there. I'm hoping this will make a big difference and actually elevate our discussion level at work, but of course that may prove to be a pipe dream.