A while ago I had a product owner come to my desk and ask me "Klaus, does everybody in the team have enough to do?". I probably looked a tad puzzled when I reflexively answered "Yes", knowing this was the answer that would satisfy her and make her go away. My affirmation told her she had done her job in ensuring that everybody's personal work-queue was full up and there wouldn't be any risk of people sitting around doing nothing anytime soon.
She left while I still was processing what that question actually meant. I came to the conclusion that Carl Sagan was wrong in saying there is no such thing as a stupid question. I think there is and I think this particular one takes the cake, maybe only beaten by "If I shave my golden retriever like a lion, will the other dogs respect him more?".
The reason why this is such an unintelligent question and how it drives bad behavior is the subject of this post. This is the sequel to a post I wrote on Agile Governance a couple of weeks ago on why jams – be it in traffic or business – happen and why as a consequence many teams really aren't nearly as productive as they could be.
Thanks to Robert Love, Bianca Legorreta, Dirk Bucka-Lassen, and Jeff Sutherland for having found slack in their busy calendars to proof read and provide valuable feedback.
Are you happy when you start up your computer and the CPU consumes 100% of its capacity? You should be, shouldn't you? If it only ran at 80%, you could have saved yourself some cash by buying a cheaper model with a less powerful processor. The same goes for the memory – we prefer it to be used to the very last byte, right? That sweet sound of a hard disk swapping memory chunks in and out is music to our ears.
Or do we like 100% utilized roads? We pay for them through taxes after all, so we should be interested in spending as little as possible on unused asphalt.
The answer to all of the above is obviously "no". We have learnt the hard way that exhausted CPUs and busy roads are not a good thing.
Why don't we apply this logic to our businesses these days, then?
Is it easy for you to find a meeting room in your organization? If not, it's probably because somebody has an agenda of maximizing the utilization of meeting rooms. It's the easiest job in the world, I might add. Declare yet another meeting room to be an office or anything else and you have just increased utilization of the remainder of meeting rooms. But now people waste even more time in trying to find a room where they can meet. Or even worse, they don't sit down together at all because they can't be bothered locating a free meeting room any more.
Sadly enough it's the same with personnel. So many businesses try to maximize utilization of individuals and teams and don't realize that they are doing themselves and everybody else a disservice.
For instance, is it easy to find a slot in your boss's calendar? Probably not – he is already in back-to-back meetings this and all of next week. He is possibly even proud of how perfectly he has utilized his work time. Of course, unplanned things need to be dealt with after hours, which means he rarely makes it to dinner with his family at home.
Other indicators of busyness
I see many teams in their daily standups reporting what they did yesterday, what they plan to do today, and that they don't have any impediments. When I ask them whether they feel they – as a team – are running at the highest imaginable speed the answer is always "no". But since they are focused on utilization, utilization is what they report on – the focus isn't on improving the process to become more efficient.
I see product owners who push the team to take in more work than they have ever proved able to do. I see ScrumMasters do the same. I even hear teams heroically proclaim "We're going to take more work in this sprint". But without changing anything about the process, how can they do more work? That's Albert Einstein's definition of insanity: "Doing the same thing over and over again and expecting different results".
What really counts
What should the focus be on, if not utilization? Output, of course (and in a second step, outcome – more on output vs. outcome in a later post). What good is it if we are busy, but don't produce anything? What matters is what comes out, not how much we put in.
The responsiveness of the computer or the number of cars that get to their destination within a reasonable time are things that it is relatively easy to measure or develop a feel for. But when it comes to people, we don't have much experience with measuring output. It seems downright counterintuitive that people, and in particular teams, would produce more if you gave them less work. How could that be?
Successfully maximizing utilization means being busy all the time. Being busy means: Too busy
- having no time to help others
- having no time to "sharpen the saw"
- being reactive rather than proactive
The first issue degrades the team's productivity in the short term. If half an hour of your time saves another team member two hours of their time, that is a good investment from the team's perspective.
The second point is about improving the process. In his book "Slack" Tom DeMarco writes “Stress kills innovation, as does busyness”. In my own experience this is a very true observation. By giving individuals and teams some slack, you give them the chance to improve their process and thereby enhance their long-term productivity. Or they invent something completely new: that's how Gmail came about – somebody created it in their slack time, or what Google calls '20% time'.
The first two points are an application of systems thinking: focusing on improving the whole instead of just parts of it.
The third point is about thinking ahead. Do you feel like you're always putting out fires, primarily working on urgent issues rather than important ones? A likely root cause for this is a maxed-out utilization.
Measure and Manage
So – why do so many businesses still keep the spotlight on staying busy?
One of the main reasons, I think, is that we still have a strong management culture of focusing on things we can measure. How often haven't we heard "If You Can't Measure It, You Can't Manage It"? McKinsey & Co. have even made this their maxim.
Unfortunately a lot of important things are really hard to measure. Measuring busyness is relatively easy – we just ask everybody if they have enough to do! The output of a team, on the other hand, is much harder to find a metric for.
Questions can be harmful
Yes, generally questions are a good thing of course. As the popular saying has it: "The only bad question is the one you didn't ask". However, we have to realize that asking the wrong questions can make people behave in the wrong way. Questions say a lot about the questioner and her motives. If your boss is regularly asking you this seemingly innocent question of "do you have enough to do?", you might very well come to the conclusion that the ultimate goal is exactly this, having enough to do.
In a way this is the observer effect, which says that just by observing the system (here through asking questions), you change the system.
Most people can probably honestly answer this question with a confident "Yes" – feeling the urge to add "I certainly have enough to do, don't worry about me" – when that is really the exact answer that should worry your boss the most.
Does Scrum solve this?
Yes and no. First of all Scrum helps highlight the problem – especially when you use story points to estimate effort. The number of story points done in a sprint, the velocity, actually is a good relative measure of your level of production (as opposed to the time taken, which is a measurement of cost). You don't get an absolute value for production and thus can't compare one team with another, but you can see how the productivity evolves within one team. This enables you to run experiments for one or more sprints and see how process change impacts your velocity.
Busyness, together with interruption of flow (see previous post), leads to jams and potentially makes both individuals and teams extremely slow.
Measure productivity instead of utilization.
Refrain from asking questions like "do you have enough to do?". Pushing work onto people is not a particularly efficient model. It's much smarter to implement a pull model, where people come and ask for more work when they need it.
Ask instead "What is your velocity and is there anything I can do to help you improve it?".
Let's focus on measuring production and stop measuring hours of wasted work.