Measure what is measurable, and make measurable what is not so. ~Galileo Galilei
In part 1 of this post, we talked about why measurement is important.
We also found that you need to have the correct intentions for making the measurements; otherwise, you will only be misled by false confidence in the numbers.
What To Measure
There really are only two important things you want to measure in your team: effectiveness and efficiency. Effectiveness is the raw output of work; when you ask for a financial report according to particular specifications, an effective tech team is able to execute the task with minimum deviation from the specs. As Peter Drucker in his book The Effective Executive famously said: effectiveness is doing the right things.
Efficiency however, can be thought of as “doing things right” (again by Peter Drucker). I like to think of it as having a multiplier effect; it does not contribute to results directly, but it multiplies whatever potential output there is. That means you want to prioritize effectiveness over efficiency; an effective but inefficient team can be optimised, but a very efficient team not doing the right things is a big waste of resources.
Effectiveness × Efficiency = Results
In most cases, you want to increase both; increasing just one or the other is like a multiplication equation. For example:
2 effectiveness × 10 efficiency = 20 results 2 effectiveness × 11 efficiency = 22 results 40 effectiveness × 10 efficiency = 400 results 4 effectiveness × 100 efficiency = 400 results
How To Measure
It is extremely seductive to think that by assigning numbers to particular activities, you get to manipulate the results — like levers on a control panel. Unfortunately, real life is not just a simple math equation (it probably is, but so far nobody had been able to completely encompass all variables).
It comes back to asking the question: why do I want to measure performance? It is not to control specific activities, but to find out where you stand as a team and discover ways to improve performance.
What can be done is to measure proxies — activities that most closely approximate the results you wanted to measure in the first place. Proxies are by definition not the exact thing you want to measure, but they can be close enough.
For example, the one of the mathematical ways of calculating Π is by Archimedes: you create triangles (whose boundaries can easily be measured) inside a circle (whose circumference is more difficult to measure) and try to approximate the the ratio from that. In other words, the more triangles he drew, the better his approximation to Π gets.
This is the same idea with measuring proxies: you won’t immediately be able to measure the exact results, but as you go along with the measurements you’ll eventually end up adding more perspectives/sides to more completely see a more holistic picture.
Ask them what have they done, what are they going to do, and if there’s anything that’s preventing them from doing it
This is a measure of effectiveness. These three questions are mostly used by Scrum methodology, but you don’t have to practice the whole of Scrum to take advantage of this approach. The three questions allows your team to:
- Commit to each other on a task
- Find out how much progress has been made on a particular area
- Improve communication with each other
- Nip possible problems in the bud
The danger with this approach is ending up with a meeting and a status update. When you start asking questions like “why a particular workset is not yet done” or “why something is taking too long,” you’re not measuring anymore (and most likely agitating your team).
These questions are not for your benefit but for the team’s; it is their measure of whether the right things are getting done, so that they know if they need to course-correct.
Ask the team to rate their feelings for the last week
This is a measure of efficiency. Most of the time we forget that we’re hiring people, not robots. They have feelings and emotions, and these are important indicators of possible issues that are hiding beneath the surface.
You might encounter some push-back, however. When I first introduced the rating of feelings to my team, there was an initial hesitation. Some team mates jokingly mocked the activity as “feelings time” and were uncomfortable sharing what they felt during the week. Some complained that it’s yet another waste of time for a meeting, and that it would only devolve into meaningless chatter.
These are all valid points, and I had to take them into consideration when I persuaded them to just try it out for a month. I promised to that we’d drop the whole activity if we found that it didn’t add any value to our team.
In the end, we did find a lot of value for the activity. We managed to:
- Unearth certain issues that were personally affecting individual team members (burn-out, work dissatisfaction, family matters)
- Address concerns on team dynamics (teamwork, communication style)
- Understand each other better
- Have a more tightly-knit and bonded team
We kept the activity, and although some team members still don’t want to talk about feelings, we just named the activity to “Sprint Score” but it’s effectively the same thing.
Ask the team to design the roadmap for your milestones
This is a measure of results. Your team are the domain experts and are the best people to tell you how long a particular feature will take. The most you can do is to guide them toward what objectives are strategic (of which you are expectedly the domain expert), but how to get there is a question you shouldn’t meddle with.
When the roadmap is drafted, have it publicly visible on a board or on a wall, and whenever an item is done, mark it down. There is something very visceral about having your objectives presented for everyone to see; this again comes back to the human and emotional aspect of teams.
Whenever something gets done and is publicly celebrated (even with just a small gesture like crossing out a line item or moving a card from “doing” to “done”), the people concerned get a rush of hormones that make them even more efficient at their work, increasing their productivity.
You might notice you have to ask a lot of things instead of objectively measuring things. This requires a very mature team that you can trust. You might need to work on trust issues first before you can even start measuring performance.
Remember that you are also part of the team, and their performance is also your performance as their leader. If you think none of these suggestions work for your team, then try this: measure your own performance using any aspect or system you think best reflects yourself (after all, you are the best person to know what you’re capable of). After a week of measurements ask yourself: do these measurements truly reflect my performance?
- Measure Results — Not Hours — to Improve Work Efficiency by Robert Pozen, senior lecturer at Harvard Business School
- The 4 Disciplines of Execution by Chris McChesney
- Metrics-Driven Engineering by Mike Brittain of Etsy