Why do people have so much trouble minimising WIP?

Have you ever done an exercise that teaches you the importance of minimising your Work in Progress? They can be quite fun, and compelling. But did you actually then go and minimise your WIP? So many teams (and people) fail to do so. Why? Because we still see people/teams as deliverers rather than value creators. As such, they have no incentive to minimise WIP.

Minimising WIP doesn’t affect throughput

If you have been involved in a Kanban process, you will probably have come across Little’s Law, that states that:

Work in Progress (WIP) = Lead time x Throughput

Where:

WIP = the amount of work you have in the system

Lead time = the time it takes for a work item to get through your system

Throughput = The amount of work the system produces per day/week/month etc

Usually we talk about how, once a team is close to full capacity, increasing WIP will only serve to increase lead time without improving throughput. Thus, it is a no-brainer to minimise WIP as we minimise lead time without having a negative effect on throughput. However, we don’t talk about 2 other aspects of this law:

If we are not at full capacity, then increasing Work in Progress will increase throughput

Yes. I said it. INCREASING WIP WILL INCREASE THROUGHPUT. Oops, now I said it again.

To illustrate this, let’s look at the extreme case. Imagine we have a team of 4, and we have a Work in Progress limit of 1 item of work. The team takes 1 week to complete this piece of work. Thus, re-arranging Little’s Law we have:

Throughput = WIP / Lead time

=1/1

= 1 item/week

However, when we look at how the team worked together, we find that 2 people actually did nothing for most of the week. The team could easily take on more work. So next week, the team took on 2 items of work. And guess what, they completed them both! So:

Throughput = 2/1

=2 items/week

Look, we have increased WIP, and it has led to an increase in throughput.

We say that minimising WIP doesn’t affect throughput. But that is only true as a team nears full capacity. If a team has a plenty of spare capacity, it can increase throughput by increasing WIP.

Increasing WIP at full capacity doesn’t necessarily affect throughput

Let’s assume (for simplicity) the team in our example hit capacity at this level of 2 work item in progress. What happens if we increase WIP to 4 now? If we can do 2 items/week, then it takes 2 weeks to complete all the work, so

WIP = Lead time * Throughput

4 = 2 weeks * 2 items/week

So increasing WIP has increased our lead time to 2 weeks, but our throughput hasn’t decreased.

Conclusion: if a team wants to maximise throughput, then it will naturally increase WIP

The above example is overly-simplified, and there are reasons why throughput might suffer as we maximise WIP. However, mathematically, increasing WIP will not damage throughput. And if we are under full capacity, we should increase WIP in order to maximise throughput, i.e. deliver more.

So if a team wants to maximise delivery, it is safer to increase WIP.

So why do we talk about minimising WIP?!

The concept of minimising WIP comes from lean production methods, which were a key foundation of the agile mindset. I’ve written before in depth about how minimising WIP makes you more agile. Essentially:

• In order to be more agile, we need to adapt quickly.
• In order to do this smoothly, we need minimal Work In Progress. Otherwise, changing course will either take a long time, or leave much work unfinished
• Minimising WIP automatically decreases lead time, i.e. it means we build things quicker
• So minimising WIP allows us to build quickly and adapt quickly. Which makes us very agile.

So if a team wants to be more agile, it should minimise WIP (without hurting throughput).

The contradiction

So there we have it:

• Minimising WIP makes you more agile
• Increasing WIP will maximise delivery

Eh?

So do we need to be more agile?

Teams are only going to be incentivised to minimise WIP if they have an incentive to be more agile. But how do we do that? And more importantly, why should we do that if it isn’t going to affect delivery?

Well, here’s the difference:

• Maximising delivery will mean that you deliver more stuff
• Being agile means you will deliver THE RIGHT STUFF

We need to be agile because our environment is very complex.

• It’s hard to understand the thing the customer wants
• It’s really hard to build that thing the customer wants
• Things change. And then the customer wants a new thing

In order to maximise value to customers, we need to iterate, to inspect and adapt, to “Build, Measure, Learn”, to be agile.

Increasing WIP will maximise delivery of stuff. But it makes it harder to learn and change course. So if we want to deliver more value to customers, we need to be more agile, and minimise WIP.

Outcomes over Output

Measuring a team’s effectiveness based on outcomes rather than outputs is the key to incentivising them to being agile and minimising WIP.

Let me define these:

• Outcomes = change in customer behaviour (e.g. increased retention)
• Output = new features (e.g. new mobile app)

If a team is focused on maximising customer outcomes, then it will need to be agile. However, if it is just focused on producing features (outputs), then all that matters is maximising delivery. As such, it will have no incentive to minimise WIP.

Historically teams were never given outcomes to focus on. Instead, they were given work to do, and measured by their outputs. In many organisations, this is still the case.

If we measure a team based on outputs, then they have no reason to minimise WIP.

If we measure a team based on outcomes, then they have every reason to minimise WIP.

Afterword: What about individuals?

I’d argue that we as individuals have the same issue. We focus on how much stuff we get done. In order to get more stuff done, we do more things at the same time. And in general, this does mean that we get more stuff done.

But it doesn’t mean we get the right stuff done. It doesn’t mean that we do our best work. There are so many reasons why minimising WIP would mean that we create more value in our lives (I’ve not even talked about context switching). I do wonder that if we were able to select and focus on the outcomes we are trying to deliver in our own lives, then we might be more incentivised to minimise the amount of stuff we do at once.

If you enjoyed reading please give some 👏 and/or leave a comment. For more stories like this check out my publication, The Agile Mindset.