2011 is a big year for the technerds, we're experimenting with a lot of new stuff, trying new ways of working, and filling up whiteboards faster than you can spit.
Before we get to 2011, let's dive back into 2010, and talk about a few process things we did that ended up playing heavily into how we decided what to work on this year. In the beginning of 2010, we started using a project tracking tool called Pivotal Tracker. Pivotal is an agile project management tool that's pretty low maintenance. It's easy to add a description of a piece of functionality you want on the site (typically called stories), estimate the effort you think it will take to complete (story points), and track progress against tasks associated with the story. As you roll along with the tool, it starts to figure out how much you can get done in a given period of time (your velocity).
The first step in optimizing anything is to measure it, which is what we did with Pivotal. As we gathered more and more data, it became clear that we were spending a lot of our time maintaining the underlying framework of our sites, rather than adding new functionality to them. Our velocity was pretty consistent, but it never felt like we were accomplishing the big things we wanted to accomplish. As an example, 2010 was supposed to be about spending time on the community, but in reality we spent a lot more time maintaining the commerce side of Threadless than we did adding new features that directly impacted our artists. That sucked. When we look back on 2010, it's one of our biggest dissapointments.
So, at the end of 2010, we started thinking about ways that we could turn that all around. We spent a lot of time with markers and whiteboards figuring out the kind of environment we wanted to work on, but needed a catalyst to start to gel everything together. Enter Cpt. skaw with his brilliant idea to use our wildly successful design model to help aid charitable organizations, Threadless Causes. It seemed like a great way to prove out new ideas we had for how to build our toys.
Causes was built entirely in Python on the Django web framework, and deployed on smaller cloud servers. Once the new stack was proven, we spent time converting chunks of core Threadless functionality to the same stack. We focused on the elements that crossed page boundries, like sessions, the cart status, and headers and footers. Which leads us to crazytown. Crazytown is a special two week project that kicks off this coming Monday (4/11/11). Tech is going to hide out in the Airstream trailers in the atrium, and pound through as many pages as we can, converting each to the new stack. As we do so, we'll roll them out to the new cloud servers.
What does that all mean to you all? Think of it as an upgrade to the buildings wiring, or plumbing. Maybe you'll notice more water pressure, or less brown outs, maybe things will be a bit tighter, but chances are - you won't see major changes. However, in the following weeks, you'll see lots of new changes, because the wiring will make it easier for us to deploy shiny new stuff. We're pretty excited for the changes, and especially excited for crazytown...we'll keep you all updated as things start to roll out.