It has been a little more than a year since we took the radical step to eliminate traditional management entirely. The goal was to extend the observations we had made from human psychology and the performance of Agile teams to the furthest extent we could imagine. It was truly an experiment: there were a handful of well-known examples to learn from, but the literature is pretty thin on the practical realities of self-organizing teams outside the software industry. And there were no peers that we knew of in eastern Iowa that had gone as far as we were intending to go. So, we rolled the dice and went for it using the best information we had at the time. Since we declared we’d be open about the experiment, it’s about time to revisit where we’re at, what’s working and what’s not. This post will just outline a few of the areas where we’ve seen success, and some where we’ve had difficulty. I’ll use the same format we use each week on each team in our retrospectives.
What’s Working Well
- Client Satisfaction is higher than it has ever been – the top three scores we’ve ever received have been in the last year.
- Communication is much faster overall.
- Most decisions are made more quickly – teams can and do make decisions all the time that required various meetings and approvals before.
- Teams learn and improve faster – they’re responsible for their pace of improvement, not when their manager tells them to. There’s no manager or organization chart to fall back on when something isn’t going right. Another factor that we think is driving the speed of improvement is that they have complete line of sight to how they deliver ultimate value to the market. On almost every team, they deliver, end-to-end, the entirety of the product or service they make. There’s no “toss it over the wall to that other department”, so they are able to make improvements along the entire value stream.
- Time dedicated to learning can be specifically prioritized against deliverable work, and it is, frequently.
- Team dynamics are much stronger – there’s more trust and candor than before. Conversations between peers, with candid feedback, happen much more frequently now.
- Politics and drama have been greatly reduced. My hypothesis is that most workplace drama comes from power imbalance or fear-based management practices.
- Pairing is an incredibly powerful tool for a workforce to realign their efforts around the demand for work rather than funneling it by specialty. We’ve seen anecdotal evidence that teams that pair more often can deliver more, faster, over time.
- Self-organizing teams are hugely liberating for a business owner. Previously, far too many decisions required CYA/escalation than they should have. I have much, much more time to think about strategic issues than I did before.
- HR and culture are strategic weapons. Previously, I think we were conflating culture with cutesy perks and events. Having a foosball table and no dress code is not the same as having a strong culture, at least in terms of driving the performance of an organization. We’re attempting to wield HR as a weapon in a way that I think is unusually aggressive, and it’s paying off.
What Needs Improvement
- Candor is still hard within some peer groups. The majority of Geonetric team members are super smart introverts and conflict-averse. Teaching teams to have direct, difficult conversations without a manager, and to do it early and often, has been harder than we thought.
- Cross-disciplinary teams are complex and hard to build. We didn’t go as far in making teams cross functional in the beginning as we maybe should have.
- The stability of a team’s membership is much more critical than we thought. We’re super nice folks, so the instinct to help out anyone and everyone is natural. But when teams “share” people, it is much harder to predict work velocity and make commitments in a sprint.
- Because most teams are now cross-functional, the people who do the same thing don’t necessarily sit near each other or work together every day like they used to, when they were in a functional department. We should have institutionalized functional sharing – communities of practice – earlier.
- While we’ve made huge strides in being incredibly transparent, there’s lots of complexity behind the scenes in financial data especially that takes time to understand. We probably underestimated the amount of time it would take to get everyone up to speed on how the accounting rules work, and why numbers come together the way they do.
- Getting pairing off the ground on teams with wildly different disciplines represented was harder than we thought. Some people felt threatened by it.
- Figuring out how to mashup a work environment without managers or performance evaluations with any sort of recognizable HR framework is much more complex and harder than we thought. Hiring, firing and compensation issues are dramatically upended in this environment (which we knew, but underestimated!). Creating alternative methods turns out to be very hard!
What We’ll Do Next / Action Items
- More pairing. Always more pairing.
- Make teams even more cross functional, and team membership more stable.
- Complete development of a unique model for compensation that is in tune with our approach.
- Continued feedback training and practice. Old habits die hard, so we’re actively working within and between teams to have better, faster communication.
- Ramp up cross-team communities of practice.
While we’ve had a few stumbles as we proceed with this experiment, the positives easily outweigh the negatives. I’m incredibly proud of this team’s willingness to try this experiment use it to build an exceptional, do revolutionary work, and “Wow!” our clients.