Like a lot of other scrum masters my teams generally work on new products. I’ve recently been scrumming a team that is rather distributed: the devs are in Costa Rica, management is here in Austin and the supporting data team is split between Austin and Dallas. There’s been great communication but by-and-large each team has kind of been doing silo work.
Last year I had a great idea… I wanted them to come together into one team working on one product so I brought up the idea of a beta. In fact I actually brought it up with the executive team and our primary customer, both of whom jumped on the idea with bright shiny eyes. Once I had everyone on board we started aiming at a beta shortly after the first of the year; our original date was Jan 8th.
So here’s the lesson I learned there: Don’t schedule major evolutions immediately after two weeks of major holidays. The two don’t mix. At all. 'Nuff said.
Our beta went live today (the guy responsible for the dev team literally just swung by and updated me about it as I was writing this!) and the change in the team is pretty palpable. Finally, we’re looking at this as a holistic entity and not three separate teams pulling on the same rope.
During the past few weeks we learned, by actual practice, not white board sketches, that our data model needed a little more love. We learned that we had infrastructure issues that no one had considered. We found out that the UI needed a few tweaks because we aren’t our customers and they don’t operate the product the same way we do.
And without that beta none of this would have surfaced until we lit this thing up live in production. You know, when PAYING customers were trying to use it. Instead we’ve got our biggest, hardest-to-satisfy customer chomping at the bit to help us design our next product. We haven’t even gotten the first formal feedback from them and already they are teaching us things about our product and process.
Was setting up this beta painful? Absolutely, but we were going to experience that pain sooner or later, at least here its under our control a bit more. Plus the team is looking at their work entirely differently and so many things that were treated as “sure its gonna work” got the chance to stand or fall on their own merits.
I think I’m approaching a new stance on new software projects. I think that going forward I will ABSOLUTELY push to get end-to-end betas (where testing occurs in a production environment if at all possible) in as many places as I can. Get the team out of the mindset of “we aren’t customer-facing yet, it doesn’t matter” and all of the ills that it can bring.