Lean Web Operations — Planning for the Unpredictable (V)

In October 2017, our managed hosting platform freistilbox had its fifth birthday. But if we’re honest, we’ve not come as far as we should have in this time. The reason is not that in a small, bootstrapped company like ours, capacity is tight. We just didn’t use the capacity we had effectively.

In this article series, I’m explaining what you we’ve learned during the last year about keeping our work in flow.

• Part 1: How we got ourselves into trouble
• Part 2: How we learned to keep work in flow
• Part 3: How we learned to make quick decisions
• Part 4: How we learned to manage interruptions
• Part 5: How we’re doing today

How we’re doing today

After all the changes we’ve made to work distribution and task management in our web operations team, how are we doing now? I’m happy to report: Much, much better! Here are the facts that confirm it:

• We keep chipping away at every project that we’ve started until it is finished.
• We’ve paid off significant parts of our technical debt.
• The time we spend on production defects per week is steadily going down. Our on-call engineers can sleep peacefully through most nights.
• Stress levels on the team have decreased significantly.
• Emergency time off is not a regular occurrence anymore.
• We’ve finally started shipping product features and improvements again.
• Customer satisfaction ratings are going up.

Since we’ve introduced the Kanban board, we have full visibility of our Work In Progress. All projects get broken down into manageable chunks. We’re able to react immediately if tasks get stuck. We’ve already gotten quite good at finishing work in time. Gone are the times when, in our weekly planning round, I was confronted with a mishmash of reasons and excuses, long after any opportunity to address them had passed. It is incredibly satisfying, both for me as the manager and the team, to watch our continuous progress.

Our project ranking algorithm has given us direction and clarity. Now we can be sure that we’re investing the little capacity we have where it yields the results we and our customers need. When it’s time to refill the WIP backlog, there’s no second-guessing what we’ll take on next; the list of important projects is already sorted and we simply pick from the top.

The introduction of the Libero role was one of the most effective changes in our history. Even though it is still quite new, our perspective on it has turned 180 degrees. First, we thought that it’s just the poor chap taking one for the team. In the meantime, we’ve realised that this role is indispensable for us delivering the service our customers pay us for.

I see the positive effects of these changes both when I talk to our engineers and to our customers. So, have we gone from gloom to glory, reached project management nirvana? No, of course not. We’re still not at the the point where we’re constantly shipping new features. The main reason is that in recent years, we had been accumulating a lot of technical debt that we have to deal with first. However, we already see that eliminating this source of constant interruptions is paving the road for rapid product improvements.

In retrospect, it has taken us almost a full year to regroup and correct mistakes of the past — there’s a thought that hurts! But, as one of my mottos goes, there is no hope for a better past. With regard to the future, I think I have reason to be optimistic. We may not be running at a steady pace, but it feels so much better to be walking in a clear direction than stumbling aimlessly through the woods. And I’m damn certain that we’ll be back in the race soon!

That concludes this article series. It’s been been the longest text I’ve written for our blog ever. If you asked me to summarise it in one line because TL;DR, I’d tell you:

Stop starting. Start finishing.