October 5, 2018
As an Agile Coach, I’ve been an advocate for co-located teams for the last 10 years. In previous posts, I’ve argued hard against the popular convention of using distributed teams. This includes why you should not use Scrum with your distributed team, and how agile transformations can be derailed by lack of co-location. After finishing the book Deep Work by Cal Newport, I’m beginning to rethink my position.
Yes, team co-location seems like a no-brainer. Anyone familiar with the “war room” concept can see the benefits of shared group knowledge and rapid, high-bandwidth communication. Nothing beats face to face communications and the authors of the Agile Manifesto included that in one of the 12 Agile principles. [See my related article on high performance teams and how co-locating teams in a War Room doubled team productivity.]
Jorgen Hesselberg provides a great example of co-located team performance in his recently published Unlocking Agility, an Insider’s Guide to Agile Enterprise Transformation. Jorgen cites an example of co-located teams that is typical based on my experience:
The co-location effort at NAVTEQ was a significant success. In less than six months after the teams had moved into their new workspaces, defects in production were down by more than 60%. The time it took to complete critical issues improved 2.5 times, and teams were delivering with more predictability and confidence.
-Jorgen Hesselberg, Unlocking Agility
In addition to the speed and richness of communications and improved quality, co-located teams often enjoy better team building, build stronger bonds, and are more likely to build the trust necessary to hold each other accountable.
However, these valuable communications come at a cost. Co-located teams can’t take advantage of lower people costs like distributed teams. Nor can they accommodate those with virtual offices, or individuals working from home.
But even more troubling to me after reading Deep Work is the inability of people on those co-located teams to have time for interruption-free thinking. Cal contends in his book that interruptions and distractions inhibit the concentration needed to solve challenging problems or create innovative solutions. This directly conflicts with the idea of co-locating teams.
The book increased my awareness of increasing technical and societal pressures to distract ourselves. Whether it is checking LinkedIn or Facebook “likes”, responding to IMs or texts, or tracking the stock market or watching Netflix, we all have a plethora of ways to interrupt ourselves or distract ourselves from meaningful work. We don’t need help from our co-workers!
Adding to the list of distractions for many people is the use of open landscape offices. This seems to be the popular move these days with Facebook, Apple and even McDonalds moving to open landscape offices. At their corporate headquarters, McDonalds has done away with dedicated desks so anybody can pretty much sit anywhere they want.
This shift to open landscape offices has been soundly criticized. Here are three recent articles that reflect the tone of the feedback:
I’ve worked in open landscape offices and I tend to agree with the criticism. It’s tough to stay focused when people are walking by or even having conversations nearby.
Advocates of co-located teams (including my former self) would argue that individuals are able to selectively tune in and tune out of those conversations. And in doing so, the overall knowledge of the team increases. Alistair Cockburn even coined the phrase “osmotic communications” for this phenomenon.
Osmotic communications seem to be great for teams working together, but less applicable when you are not sitting with your team. Or, when the work you are doing is individual and not team-based. When people need to focus on individual heads down work, interruptions and conversations can kill productivity and stifle creativity.
Going back to Deep Work, Newport does push for individual isolation for productivity. However, he also recognizes the value of serendipitous meetings and collaboration. He cites the ability to interact with another to co-create, particularly at whiteboards, that sounded very much like what we expect from agile teams.
For some types of problems, working with someone else at the proverbial shared whiteboard can push you deeper than if you were working alone. The presence of the other party waiting for your next insight—be it someone physically in the same room or collaborating with you virtually—can short-circuit the natural instinct to avoid depth…By working side by side with someone on a problem, you can push each other toward deeper levels of depth, and therefore toward the generation of more and more valuable output as compared to working alone. When it comes to deep work, in other words, consider the use of collaboration when appropriate, as it can push your results to a new level.
– Cal Newport, Deep Work
Jorgen Hesselberg described similar experiments he conducted at Intel with similar results.
For instance, Intel has built co-working spaces throughout its campuses worldwide so that people can just as easily work in semipublic spaces bustling with colleagues as they can in private, two-person task rooms. Having options allows for flexibility. People can enjoy the buzz of a creative space when needed and take advantage of solitude when deep concentration is more appropriate. The key element of this workplace design strategy is choice: employees get to choose which environment they want to work in based on their personal work needs and the specific nature of the work itself. One size does not fit all.
-Jorgen Hesselberg, Unlocking Agility
It seems that there is a blend of individual heads down work as well as shared work that creates optimal productivity. How do we design spaces that allow for that? And do distributed teams have an advantage in that team members can move from isolated to connected and back seamlessly?
The Deep Work book shook my beliefs about the value of co-located teams. I know the economic reasons many people choose to use distributed teams. What are you seeing? How does team distribution impact team performance? Have I been wrong about co-located agile teams?