“If you can not measure it, you can not improve it.” – Lord Kelvin

My recent experience using The Eatery got me thinking about how much having the right measures helps you track and influence performance. Data is awesome and measures are great — nothing focuses the mind more than a graphical representation of how broken something is.

Unfortunately mistakes are often made:

• Picking what CAN be measured, rather than what SHOULD be measured. For example, airlines are reputed to determine an “on time” departure based on whether the plane left the stand on time — not when it actually took off. As a customer who has spent a lot of time sitting on the runway, this is pretty annoying!
• Focusing exclusively on the measure, rather than what it tells you about performance. We’ve all seen a team get obsessed about getting a particular measure to “green” or “100%”, often not really thinking about what the impact will be of doing so. Sometimes it’s fine to have 87% — and the incremental work to get to 100% is time & effort better spent on something else.
• Losing track of the bigger picture. My personal peeve here is measuring whether everyone in the organisation has an annual review (because that’s something easy to measure) rather than whether they actually make a difference — the folks over at Sonar 6 have covered this in one of their colour papers on performance management very effectively.

All this can lead to what one of my colleagues refers to as the Watermelon Phenomenon — when something is green on the outside, but bright red on the inside. Effectively, your measures are telling you everything is fantastic, but your users/customers/employees are telling you it is B R O K E N.

So, what to do?

1. Think about what REALLY tells you whether your business/process/organisation is healthy. Is it sales? Is it customer service? Retention rates? Happiness? How on earth do you measure HAPPINESS? Don’t worry about that right now. Just think about what you think are your key indicators of success/health, in an ideal world.
2. Now, what could show you AFTER the fact how you are doing? These are called “outcome measures” or “blackbox measures” and are typically hard numbers — sales, profit, retention, etc. The disadvantage is they are often broad and reported after the fact — but they are still important.
3. Then, think about what your early warning measures might be. In people management, how many people you are attracting (recruitment) and keeping happy & engaged (retaining) are your outcome measures … but how do you tell BEFORE someone goes as far as leaving that you have a problem? Is it an employee health survey, the team’s feelings about their manager, how much your competitors are paying? Working this out and then finding a way to measure it without losing track of the end result, is probably the hardest challenge.

What’s the worst measure you’ve ever seen, in terms of the behaviour it drove?