At Atlassian, we are big believers in the concept of 20% time. We’ve being doing it for more than 4 years, but it hasn’t been without challenges. Probably the biggest surprise for us was the fact that as a company, we actually only used 1.1% of the 20% time allocated. How come?
There are a few reasons for this:
- Scheduling time – We ship releases all the time… so there is never a good time to take 20% time
- Workload – Letting other people pick up the slack while one works on their pet-project is not cool
- Dependencies – People depend on each other for their work. Being offline for a day or two can potentially block progress on your team mates work
- Visibility – Even though we are tracking all 20% time projects with Jira, visibility can be an issue
Let’s package it up
The #1 problem for us has been finding the right time to take 20% time. People didn’t want it to conflict with their regular work or inflict an additional work load on their colleagues. That’s why we’ve been trying something a little bit different with a few teams… introducing “Innovation Week”.
- Innovation week is held at the beginning of each release. That way, potentially shippable features can be included in and polished during the release.
- The Product Manager communicates the product strategy to the team and points out important focus areas. Everybody wants the product to succeed and knowing what’s important for the product right now can inspire good ideas. The important part is that the product manager doesn’t suggest concrete features or specs… let the engineers be creative and be surprised by the results.
- Similar to Ship It days, people will have to submit a shipment order explaining what they will be working on and how it will contribute to the products (or company) success.
- Forming teams is encouraged. Since everyone takes their 20% time together, it’s much easier to work in a group.
- Designers and product managers should be available to help and answer questions.
- The goal is to present a shippable feature or a prototype of something more complex by the end of the week.
Innovation week has been really successful. In total we’ve shipped 12 features and improvements in 2 innovation weeks alone. These shipments ranged from reducing the memory footprint during a database migration, over switching to less for smaller and easier to maintain stylesheets, to major features such as the Image Diffs and File Search that we’ve shipped with Stash 1.1 and 1.2 respectively.
Furthermore, we’ve had a whole bunch of spikes which, although they haven’t shipped with the product, gave us a pretty good idea about the features, how to best implement them and how long it will approximately take us to get it done.
Best of all, the team is loving it. It feels more like a week long Ship It where everybody works with the same goal, to present something amazing by the end of the week.