Recently, I was thinking of some of the problems I had in the early days of composing and noticed an interesting paradigm between how modern companies develop software and the practices I’d adopted when creating music for clients.

Years ago, software developers did their work in a straight line. A client wanted a system and you started building it from the ground up, not stopping until it was near completion. Then, you showed it to the client, tested it and usually had to go back and change a ton of things since the client wasn’t really sure what they were asking for or you misinterpreted it.

To resolve this problem the Agile Scrum process was created to solve some important problems with the traditional method. This idea was developed in the 80’s but in the early 2000’s it really started to become adopted on a large scale by many companies. I’m going to keep it very high-level here (developers aren’t my main audience, I think?). The aim was to cut down in the vicious cycle of developing done without any communication to the end user and then going back and redoing it when the end user didn’t like it. It also addresses how teams plan and deliver things.

One of the most important pieces is incremental delivery. This is something I’ve adopted into my music composition process when working for a third party. Much like the traditional developer, my early soundtracks were seen and heard only by me until completion. Then, when the client got the finished product, they usually said, “this part is too long”, “we don’t like this instrument”, etc.

In the Scrum methodology, you have a “sprint” which is a set of time to complete a unit of work. A unit of work is something you can show the client at the end of the sprint. This way the client knows what on earth you’re billing him for and you also can be more sure the client will like the finished product. This is exactly what I do with a soundtrack now. Here’s my sprints:

1. Initial sound of the album – I create a few 30 second demos, not properly mixed, some default oZone mastering on it and send them to the client. I usually make them loops so they can try them in game. From here, we determine if they like where things are going or if I’ve misinterpreted their directions and suggested feeling.

2. Short Drafts of Each Track – If the soundtrack will be 10 songs in total, I make 10 30-45 second looping tracks displaying how I envision each track based on their direction. I usually try to include at least 2 dynamic shifts in this short period of time so that they can envision how the final version would sound. This could be more than one sprint depending on the amount of tracks needed.

3. Final Drafts – Once the client is happy with the short drafts, I develop all of these short unmixed and unmastered demos into complete tracks based on the duration the client has requested. At this point they have their last chance to weigh in on instrument changes, tempos and other factors that play into the mixing and mastering phase. This section may be many sprints if it’s a very large job.

4. Mixing and Mastering – Now we do our mixing and mastering and create finished products. We know we are safe doing this now since the client has confirmed everything up until now, what we send them in the end will only be slightly different from what they heard in the “Final Drafts” stage.

5. Final Preview and Delivery – Once we have them all mixed, I send an archive file of mp3 versions for the client to listen to one last time. Then, if they’re satisfied, I send them the formats they’ve requested, raw WAV files and any stems if they needed them.

The great thing is that this is a great way to break up a large contract into multiple payments. Any completion of a sprint is a good place for an incremental payment.

Here’s a few tips for working with this mindset:

1. Save everything! – Don’t think a song is done when you’re done with it. Save each draft you make. If you make 3 versions with a different lead instrument, save three projects so you wont have to remember what it is later. Once, when I used a hardware synth to make an early version, I didn’t write down the settings, and I could never quite capture the same sound later on for the final. Keep track of everything.

2. Be Organized – It’s totally worth the time to move things into folders, save backups and make spreadsheets of where things are. If you don’t, you will forget something at some point.

3. Communicate – Make sure you’re client understands this process. One important thing is that they realize your first drafts are not final projects. They need to remember that it’s just to get a mood or feeling to start with.

What are your methods for composing?