Whew. What a crazy project this new LockerPulse has been. Remember when I posted in February that we were half done after about 6 weeks of work? Well, that 3 month project has turned in to a 6 month project. Generally I’m pretty spot-on with my estimates when I plan these types of large development projects out. What slowed us up this time?
Without a doubt, running Detailed Image is reason #1. We’ve grown significantly since our last large development project, the original launch of LockerPulse in 2010. I severely underestimated how much time Mike and I have to spend daily on managing/training employees, communication with our team, customer service, accounting, marketing, etc. The long stretches of development time without interruption are still there, they’re just harder to come by on a consistent basis.
The second reason is that we’re approaching this project a little differently. We’re taking our best shot right here. It’s our home run swing. As I mentioned in that post back in February, we’ve got what we think is enough data to justify taking that one gigantic home run swing with this project. With that, the level of detail we’re paying attention to is magnified. Instead of cutting every feasible corner to launch as fast as possible (something I’m 100% in favor of for most projects), we’re allowing ourselves to spend an extra few days to get something right if that’s what it takes. The fact that we don’t need LockerPulse to bring in any revenue any time soon for us to survive works to our advantage here (although one can see how it could work as a disadvantage if you’re not careful).
Lastly, we’ve had some bad luck. For instance, our data provider for rosters, schedules, and standings shut down their API last week. We kind of saw this coming – the service was increasingly bad – so there was a backup plan in place that I got working on as soon as they made an announcement. In the end it’s for the better, but it easily cost us ~2 weeks and stressed me out quite a bit. As of a few weeks ago we were hoping to launch on 5/12/12, the two year anniversary of our original launch, but now it’s looking like sometime around Memorial Day. Such is the risk any time you rely on API’s.
As it’s dragged on (well, dragged on by my standards at least) I’ve been plenty stressed and frustrated. I’ve worked far too many hours for the work/life balance that I want. The good news is that the serious development is almost completely done. The other good news is that it came out awesome. I absolutely love using it. Seemingly all of the attention to detail has been worth it. Now comes the fun part: launching so that the rest of the world can use it!