Distributed Agile teams may not be the norm, but they’re here to stay and their numbers are steadily growing. In my own experience, they can work pretty well and become quite productive (see my prior post), although there always seems to be a “remote penalty” compared to fully co-located teams, basically the inevitable cost and friction introduced by not being in the same place.
Is it possible to boost such a team’s performance further and reap benefits similar to co-located teams? Can you “virtually co-locate”?
A recent experiment has convinced me that the answer to these questions is “Yes”!
The Starting Point
I was working with a “regular”-sized Kanban team that had already worked together for a while and found its stride. Members were located from the US West Coast, the Mid West and Florida to South America with a maximum time difference of 4 hours. Things were going pretty well: We were conducting the typical Agile ceremonies via Zoom sessions, used Azure DevOps to manage work, collaborated via Slack, and had working sessions and meetings as needed.
The team had been working on a new module of a SaaS application and we were faced with the challenge of delivering the MVP within about 5 weeks, which was doable, but not easy. We discussed how to work together differently to intensify our efforts and increase our chances of delivering on time.
We had the option of traveling to a central location and work together in the same office for a few weeks, which obviously carried not insignificant costs such as travel time and expenses, logistical challenges including visas, and being away from our families. While we left that option on the table for later, the team decided to try out a different set of working agreements first.
What we tried was in essence equivalent to “virtual co-location”: We agreed on an 8 hour time period during our central time zone every day when we would work together with all – or at least most – team members present. During this time we would use and keep open a primary Zoom session with a second Zoom as optional break-out. The team members would all join the Zoom session, stay connected, and – if needed – some individuals could peel off and jump on the secondary Zoom for a sidebar or one-off conversation. This arrangement was not easy for all logistically as it meant starting at 6 AM PST every day for the US West Coast while it meant starting and ending late in South America.
The benefits were many:
- Pairing or even swarming on key deliverables became the norm.
- Frequent and extended high-bandwidth communications.
- Osmotic communication: Those not directly involved in the primary conversation taking place would still overhear what was being discussed and chime in as needed to ask questions or provide other inputs. (This is similar to what teams sitting in the same room experience.)
- Increased camaraderie among the team members.
- Questions and blockers were addressed almost immediately.
- Almost no scheduled meetings any more (apart from the typical ceremonies) unless we needed input from people outside the team, such as advice from our UX advisor, etc.
Apart from early start time and late days, there were not too many downsides, maybe apart from the challenge of figuring out when everyone would squeeze in lunch.
Our new M. O. of virtual co-location felt intense and seemed to accelerate our ability to get things done. We were able to meet our deadline for our MVP and deploy as planned, which was a big accomplishment for the team. While we brought up the questions of traveling to a central location a few more times as an available option, the team clearly preferred to work in this new manner.
Looking back at our team metrics, I was shocked: Our average velocity increased by 90% and our throughput increased by 95%! I knew the metrics were going to be better, but I didn’t expect this type of drastic jump. There were certainly other factors involved, such as people putting in extra effort towards a final deliverable, but we also had some people being out due to holidays, etc. I’m not a “sensationalist” and won’t claim things like “Do twice the work in half time with Scrum!”, but suffice it to say, this change of working arrangements had a significant impact on the team’s performance!
Made to Stick?
So with the MVP out of the way, what happened next? Did this new way of working “stick”? Well, life’s a journey and so is every team’s evolution. We have learned a lot from our experience, formed deeper bonds, and developed new habits of problem solving and pairing. While for now we discontinued the standing Zoom bridge and loosened the requirement for everyone to be online at the same time, I’m hopeful that we’ve built new patterns that will persists, even if more informally.
If & when we face the next major challenge as a team or we find ourselves feeling disconnected, I can definitely see us reverting back to virtual colocation, even if only for some time or in some modified way if necessary.
For now my theory is that this new way of working was the productivity boost we needed at the time, but that there was a certain amount of overhead that made it not fully sustainable in the long-term.
Still, this experience has shown that distributed teams have ways to achieve high performance, comparable to that of co-located teams.
Are you ready to try “virtual co-location”?