Teams are better than individuals
Accelerate delivery by working as a team

Teams are better than individuals

I have been fortunate to be part of some amazing software product development teams, so I know well the feeling of acceleration as a good team gets faster and faster at delivering. Inspired by the wonderfully simple way in which Henrik Kniberg demonstrated the benefits of limiting work in progress in his 2016 AgileByExample keynote, Focus (or Stop Starting, Start Finishing), I wondered if I could use the same approach to demonstrate the benefits of working as a team. It turned out I could, and the results are striking!

Measuring the benefits

In order to demonstrate benefits, we first need a way to measure them. In my projects, I'm interested in delivering valuable stuff. That's hard to measure, so I'll use time to market as a proxy. Having a shorter time to market doesn't guarantee delivering more valuable stuff, but it does mean a) we'll deliver more stuff, and b) we'll learn sooner if the stuff was valuable, allowing us to select the most valuable stuff to work on next.

I'll also look at the impact on total input time, or the total time spent working on tasks. I've included it because it's interesting, but be careful. Optimising for inputs can (will?) decrease the amount of valuable stuff you are able to deliver (that's a different article, but research 'overutilization' if that's a new concept to you).

Notation

Full credit to Henrik Kniberg for the notation, which is the same as used by him in the presentation referenced in the opening paragraph.

Each different colour corresponds to a 'project'. This could represent a project, an epic, or some other package of work that is self-contained, and ready to be delivered once completed.

The rectangles showing smileys at different stages of completion correspond to 'tasks'. These could represent a user story, technical task, work item, or some other discrete piece of work that is required to complete the project.

Elapsed time is on the horizontal axis (wider things take longer), and the different rows represent different people available to work on stuff.

1. Team vs. individual - what's the difference?

First let's compare the basic case of working on projects as individuals, versus working on projects together as a team.

Teams spend the same amount of time on tasks, but have shorter time to market

As individuals, person A works on each task in project dark-blue until it is completed; person B does the same on project mid-blue, and person C does the same on project light-blue. Conversely as a team, everyone works on project dark-blue until it is completed, then they move on to project mid-blue, then project light-blue.

You can see that total input time is the same in each case, but time to market is shorter when working as a team. Project light-blue is delivered at the same time as working individually, but dark-blue and mid-blue are delivered sooner. Working as a team wins in this simple case.

2. Does coordination slow us down?

One of the challenges I've heard to working as a team is that there is a need for coordination that you don't have when working individually. This takes time, and therefore it's less efficient to work as team. I would argue that there are lots of additional benefits that coordination and collaboration bring, but let's put that aside and look at it at face value.

Teams spend longer on coordination, but still have shorter time to market

You can see that there is now a gap between each project, representing the extra time taken for coordination. The total input time is slightly longer. However the time to market for each project is still shorter when working as a team. Project light-blue is delivered slightly later than when working individually, but dark-blue and mid-blue are delivered sooner. Using our measure of time to market, working as a team wins. However, you might be a little uncomfortable that project light-blue is delivered later. Hold that thought as we dig a little deeper.

3. Faster learning = better use of time

One of the reasons I gave earlier for using time to market as a proxy for delivering valuable stuff, is that a faster time to market lets you learn faster, and make earlier decisions about what is valuable.

Teams learn faster, so do less work to achieve the same outcomes

Here you can see the team measured and learned earlier. With those learnings, they decided that project light-blue wasn't actually the most valuable thing to work on next, so they reprioritised, and worked on something else instead. However the individuals didn't have the opportunity to learn this until the end, when they'd already wasted time completing project light-blue.

Not only was the time to market shorter per project when working as a team, the total input time was also shorter, because the team was able to scrap a whole project. This freed the team up to work on the next project sooner. Working as team wins again, and now we're starting to see a clear acceleration over what can be achieved working individually.

4. Won't less experienced team members slow us down?

Here's another challenge I've heard. Some people take longer to get stuff done, so if we work as a team, everything will go more slowly. Let's take a look, and see if that's the case.

People work at different rates, working as a team decreases time to market

You can see that individually, person A works faster than average, person B works at average rate, and person C is slower than average. When they work as a team, the members can share the work by splitting tasks into smaller pieces, or working together.

Time to market is smaller as a team than for the fastest individual working alone. When compared to the slowest individual, it's much faster. Total input time is less for the team, because the faster team members are able to take a little more of the work, and / or help slower members to complete tasks faster. Working as a team wins again.

5. Combining skills to knock it out of the park

In a team, you have some members who are slower, and some who are faster. But you also have a mix of different skills. What happens when a team combines those skills effectively?

Combining skills in a team gives striking improvements in time to market

You can see that person A is faster than average at starting smileys, but slower than average at finishing them. Person B is faster at the middle section, and person C is faster at finishing.

When working as a team, each person can bring their skills to the project, and knock it out of the park compared to working individually. The acceleration in time to market is the most striking we've seen so far, and the total input time is much reduced. This is a resounding win for working as a team rather than working individually.

Conclusion

Using simple notation, I've demonstrated that working as a team is more effective than working as a group of individuals. In some cases, total input time might be slightly longer as a team, but the time to market is reduced, meaning we're able to deliver more valuable stuff. When you take into account the differences in work rate and skills of team members, and the ability to learn what's valuable sooner, the team soon accelerates away from what is capable by a set of individuals.

The examples were simplified, but what they show matched my own experiences of working in good teams. There are also other benefits to working in a team that I haven't shown above, including:

  • Increased knowledge sharing between team members.
  • Everyone can work on any project or bug, reducing bottlenecks.
  • Teams don't need to rely on 'star players' to function well, but star players still shine, by multiplying everyone's efforts, not just their own.
  • Less-experienced team members learn faster.
  • Context switching / cognitive load for helping team members is reduced, because everyone is working on the same project.
  • Bugs and ad hoc work are easier to schedule.
  • Work in progress across the team is reduced, meaning less to manage.

The whole really is more than the sum of its parts.



Some of my other articles you might be interested in:

Tim Benjamin

Techie, storyteller, disruptor, creator, CTO, NED, and composer.

2 年

Nice one! :-) In your "as a team" approach, doesn't that fall into "mythical man-month" territory? (i.e. if it takes 1 person 3 days to do X, then it will take 3 people 1 day - which I remember I once illustrated with "if it takes 1 woman 9 months to have a baby, then it will take 9 women 1 month...") On the other hand, with your "working as individuals", that looks like a single piece flow approach (each worker completes every step of the process, manufacturing one complete item at a time), which is measurably very efficient, and often more efficient than a production line approach (each worker does just one step of the process / each worker does all of step 1, then all of step 2, manufacturing many complete items at once). With single piece flow, value is delivered faster (but in smaller quantities) than a production line. But that may not be what you're saying here. Surely it only works if your tasks can be broken down into something a team can frictionlessly work on together - in reality isn't the added friction (e.g. communication) often what gets in the way?

回复

要查看或添加评论,请登录

Gabrielle Earnshaw的更多文章

社区洞察

其他会员也浏览了