Scrum for Rock Bands. How to measure progress when recording

The band splits

When rock bands go into a recording studio, they don’t always get the job done the way they want to. It tends to go as follows:

• They start off with a lot of energy
• The stakes are high, so they invest a lot of energy in all the details. They want the best result possible.
• They get stuck on those small details, and lose a lot of time.
• At the end, they notice a total lack of time to record everything they need. Even the basic tracks have to be rushed.
• Quality decreases. The band starts to argue. The band splits. (This last point is exaggerated, but it’s no exception.)

So what is happening here? There is a lack of focus. There is no idea at any point on how much work has been done, and still has to be done. There is no measurement of progress. At the end everybody panics, resulting in poor quality.

These issues are also common in software development projects. That’s why I used scrum when recording an EP with one of my bands.

Building a backlog

At the start of the recording session, the band sat together. Every instrument that had to be recorded for every song was written down on a post-it. The backlog contained the following types of cards for every song:

• pilots: tracks we used as a reference 
• drums
• bass guitar
• guitars
• lead vocals
• backing vocals
• keyboards and special effects

Estimating the weight: planning poker

Next, we did a planning poker session. As I didn’t have my planning poker cards with me, we used our hands to give estimates.

(Please note that only two of the band members have any affinity with software development. The others were novels to scrum. But that didn’t cause any problems. The simplicity of these techniques allows to use them with anyone who shows involvement and wants the project to succeed.)

We used the recording of the pilots as the reference for our estimates. We gave it 3 points. All the other cards had to be estimated by everyone simultaneously, in relation to the “pilot card”. I explicitly talked about complexity instead of time, when rewarding points to cards.

Some examples:

• Lead vocals song 1: 5 points
• Bass guitar song 2: 2 points
• Backing vocals song 1: 4 points
• …

The total count of all the cards in the backlog was 110 points.

Prioritizing the backlog and scope creep

Next step was to put the post-its on the windows (the walls didn’t stick very well), and to prioritize them. Drums and bass were put at the top, followed by guitars, lead vocals,… . We used 3 windows: one for “to do”, one for “ongoing” and one for “done”.

At this moment our lead singer stated there was a lack of space for improvising. So we decided to create an extra card called “Inspiration”, to give it 20 points, and to put it at the bottom of the backlog. This way the time-consuming improvisation part was given a name and deprioritized in advance in relation to all the other tracks that had to be recorded.

(Project managers may notice the analogy with scope creep, where projects fail because of the constant addition of new features to the scope. The project runs overtime and the quality decreases.)

It was great to see that scope creep was defeated in advance, by giving it a name (“inspiration”) and by giving it a weight – 20 points – in relationship to the other tasks.

Measuring progress: the burndown chart

As I wrote earlier on, the most difficult thing when recording is measuring whether you’re on track. Are we working fast enough? Are we working too fast? Will we finish in time?

We had a fixed timeframe of 5 days (1 sprint) to record everything for the 4 songs (130 points). So I made the following Excel chart in Google Drive, where the ideal progress was 26 points per day.(130 divided by 5).

Next, I made a burndown chart, plotting the days on the X-axis, and the points that are burned on the Y-axis. The red line indicates the ideal progress.

The result

The first day, we were still on track. Only the second and the third day our velocity decreased. Mainly because of the lead vocals that weren’t going as smooth as foreseen. But as you can see on the graphic below, we immediately noticed that we were putting too much time in those vocals after day three. Thus we speeded things up a little bit, going into a hyper productive state the fourth day.

We burned 24 points the first day, 17 points the second and 19 the third. On the hyper productive fourth day we managed to burn 46 points. The final day we had only 24 points left. 4 for a keyboard track, and the remaining 20 points for inspiration.

Learnings

• Scrum can be used for all kinds of projects (not only software development)
• Your team can consist of all kinds of people. They don’t have to be hardcore agilists with a fetish for charts and post-its (like me). As long as they’re committed and motivated, they can perform in a scrum environment.
• Give scope creep a name, and talk about it early in the process.
• Don’t be afraid to experiment with scrum or other techniques. Even in a “rock’n’roll environment” they can prove their value.

P.S.: If you want to hear the results of the recordings, follow Lapaz on Facebook. We’ll release the EP somewhere in March.

P.P.S.: Full blown scrum consists not only of artefacts (as the backlog and the burndown), but also of roles (team, product owner, scrum master) and timeboxes (planning, daily scrum, review & retrospective). Read all about it in What is Scrum?