In knowledge work, we love metrics. Everyone wants to capture them, managers want reports via shiny dashboards. We say we need them to make decisions or better yet, influence behavior. They are the basis of target setting. If you have no targets, it really doesn’t make sense to have metrics. Yet your targets may be hurting you in ways that you cannot even imagine. As Eli Goldratt said:
Tell me how you measure me and I will tell you how I will behave.
Whenever we give people targets for them to meet, we are invariably sharing with them their performance unit of measure. When they understand what is being used to measure their performance, behaviors are modified as a result but the wrong behavioral modifications can adversely affect an organizations ability to succeed. In fact, the wrong behaviors can be devastating. There is so much information on this that its quite disconcerting that managers are still being tricked into doing the wrong thing. Check out the work of Deming, Ackoff, Goldratt and Seddon for starters.
If you’re in the software development space, this next story may resonate with you. We had a drive where we wanted to get our bug count down and so teams (and individuals) were tasked with closing a certain number of bugs daily. The metric was “bugs closed per day”. At the end of the exercise however, we determined that although we met the goal of bugs to be closed, we had introduced a brand new set of bugs that were just as bad (or in other cases worse) than the ones that had been previously closed. Because we were so focused on closing bugs, we did not pay attention to new ones we were introducing.
But the example above does not just apply to product development. We find targets (and the related metric) that alter behavior in all walks of life. Asking doctors to see a certain number of patients a day, telling police officers to give a certain number of tickets a month, setting targets for auditors to perform a certain number of audits a day, challenging developers to bang out a certain number of lines of code an hour inevitably leads to these individuals modifying their behaviors as a result of the target.
Deming’s System of Profound Knowledge (SoPK) challenges leaders to have some understanding of psychology. We need to know what influences peoples behavior and motivates them. It’s been said that the best motivator is the manager who is not demotivating. Yet we find many of targets that we set end up demotivating because they take the joy out of the work. This happens because of one of the following occurs:
- Targets are met but desired outcome is not. Like in the example above, the bug count went down but quality did not improve.
- Targets are not met, let the blame games begin. When we cannot meet our targets, our natural instinct is to look for someone else to blame. I’m sure the “system does not work” resonates with a few of my product development friends. Someone else is always to blame.
- Targets are flat out unreasonable. There is no way they could be meet under the current circumstances and setting them was unrealistic.
So are targets (and metrics) completely evil? I don’t think so. But the wrong target can be completely debilitating. So here are some guidelines on setting targets (if you must).
- Choose the right target. The outcome that is desired has to be explicit and made known to everyone. In sports, the outcome for any team is to win their league. For a software development team, it may be to have a quality product. For a police department it may be to reduce the number of accidents on the freeway. There are always outcomes that for which that entity exists and must meet continue to meet for the entity to remain viable. The relationship between a target and desired outcomes should be direct and not inferred. If a certain amount of money must be brought in weekly for a business to stay afloat, that number should be available to everyone who can influence it.
- Engage the team in target setting. When the team fully understands what the outcome needs to be, they will be able to provide suggestions regarding what the targets should be given the constraints. When they are engaged, then they are invested and then there really cannot be excuses. Don’t insult your workers by hiding this information from them. If you think they are not capable of processing it, then you need to leave because you hired them in the first place. But if you hired good people, trust them to help you set the proper targets.
- Revise targets when needed. No condition is permanent. As the landscape changes and the organization learns and evolves, targets may need to be revised (or even changed). Don’t fall in love with your targets. They are not an end in themselves.
Take a look at the targets and metrics your organization has created. Are they meaningful? Are they encouraging the right behavior? Are they motivating the workers? Are they creating organizational success? If they are not, toss them out. It’s time to revisit the drawing board.
I’ll leave you with this….