Two years ago we shared our worfklow and work principles at Colloq. Since then, a lot changed, we experimented quite a bit with various approaches and now I want to describe our current new state of “The tools we use to stay afloat”.
Strict Scrum was a bad idea
Back in the days, we decided that we want to work with Scrum as a project management framework and we’ve built our workflow around that. I, as a freshly baked certified Scrum Master tried to implement a very strict, complete process with a lot of requirements, boundaries and strict limits and goals. I wanted us to work how I experienced it in one of my client’s team and also exactly how it’s written in the books and teached everywhere. That means two week sprints, estimations with Story Points (using Fibonacci numbers), plannings, standups, reviews, retrospectives, and also by analyzing and embracing the net value of our team, usually done by the amount of Story Points achieved in one Sprint (also known as “Velocity”).
It all was ambitious to for a team like ours as we only work part-time for Colloq and most of the time during the week for our individual clients. And we failed.
What does failure mean
When I say we failed in following our own process, it means that we nearly never reached our planned Sprint goals and if we did, we weren’t happy because we knew we had estimated the Sprint with very low numbers that wouldn’t make us happy even if we had done everything. If that happens for a couple of sprints that’s not necessarily bad but despite trying to improve the process through extensive retrospectives and improvement time boxes or changing our tooling we failed constantly for more than a couple of months.
We didn’t realise that the half year of constant, perceived failure left its traces in our minds
It doesn’t mean we did nothing, quite the contrary. We built quite some nice features, refactored a lot. But this the process we wanted to follow didn’t work out for how we could do our work. Towards the end of 2018 “we gave up” and decided to give ourselves the freedom of no process until the end of the year. We hoped that every single of us would be motivated enough and wanted to follow their own vision and build the features for it.
However, we didn’t realise that the half year of constant, perceived failure left its traces in our minds. We just weren’t motivated enough and instead of building features we hardly did do much more than software updates, dependency updates or bugfixing. Which all is required, too, but it’s not helping to move the platform in the direction we want it to be.
So in 2019 we re-applied a bit of structure and tried to motivate ourselves. Here’s what we do now:
We sort of follow a Scrumban workflow now, where we do a review, retrospetcive and plan the upcoming tasks for the next week on one day. None of the tickets is estimated though — we simply don’t have the time for this and don’t have the urgent feeling of needing that at the moment. If we have a big story coming up, we try to task it out together and roughly estimate it but we finally don’t put any numbers on tickets anymore. For us, this works fine and is a relief as we save ourselves at least two hours of estimating tasks each week.
If we don’t finish our tasks in this week, it still doesn’t feel great but we don’t have a huge pressure on ourselves as there’s no strict Sprint goal. We just try to identify the blockers, how we could do it faster or smaller in order to finish it in a shorter time period.
Instead of strictly following our roadmap that we’ve come up with after countless meeting-hours, we now let each of us decide what we want to focus on next. This means everyone in the team has the power to do whatever they feel needs to be done now. This is a big relief as previously we could never reach a conclusion everyone agreed on what the most important thing would be now.
Our retrospectives are the most important meetings we have. They always were and I strongly believe that without them we wouldn’t exist anymore today. This is in line with my experience from a variety of teams I work(ed) on: Those who have a good process of doing retrospectives and actually trying to solve issues that come up during the meetings perform way better and the people in the team are much more happy than those who don’t actually experiment with changes when a problem is coming up.
We said it two years ago and while in the meantime we once tried Trello in addition to Github for exactly one week, we can still say: We’re still managing Colloq entirely in Github. We use Projects (now only one project in our organization, no individuals per repository) and we use Team Discussions for idea explorations and feature discussions.
We still use Slack as it’s an easy way to communicate. Ideally, we think we’d move to another tool that is supporting asynchronous work more and might also be self-hosted but in the meantime we’re still happy that we can use Slack.
For our calls we switched to FaceTime Group calls recently and it saves us $15 each month we paid for appear.in. We always liked appear.in and it’s why we supported it with our money but having a natively integrated app like FaceTime is just more convenient and actually free. We also like zoom.us as video conferencing tool but don’t have the need currently.
All in all, I think we’ve made good progress during the past two years and learned even more during the countless workflow and project management strategy changes we did.
Overstructuring a process is bad, while no structure (…) is demotivating individuals
Mostly, we learned that overstructuring a process is bad, while no structure gets nothing done and is demotivating individuals. We learned that it’s okay to not give 100% all the time but instead it’s necessary to have calmer times as well.
We’re constantly trying to match that new, unconventioanl expectation now.
We’re all just humans who have a daily job, who need to earn money first and then can take care of projects like Colloq. We’re not taking any investments for Colloq and thus aren’t making any money besides covering our costs for running the business yet. Until we’re more profitable, we need to make compromises — whether we like that or not. Making the best out of that is what matters and we’re constantly trying to match that new, unconventioanl expectation now.