[duey]

I've witnessed this first hand - we recently started rolling out measured goals in our warehouses. This involved staff getting set daily goal counts for certain tasks (products shipped, products put on shelves etc.). As soon as we rolled out metrics everything that wasn't measured became a low priority. For example, we were not measuring cycle counts (stock take) so immediately the number of cycle counts plummeted, which increased our out-of-stock errors. We had instances of staff hoarding incoming shipments from suppliers at their workstations in order to get higher rankings.

Overall the warehouses went from being generally efficient to extreme performers on measured metrics. I think metrics can be clearly powerful, but you have to be very careful about what metrics you choose to implement.

[rpwilcox]

I really wish this was taught more in business school: that metrics/productivity is subject to the Observer Effect (http://en.wikipedia.org/wiki/Observer_effect_%28physics%29) too: that the act of observing for metrics will change the underlaying system.

If underperforming on your metrics will cause some pain, humans in general want to avoid pain, and you ARE employing smart people who are paid to solve problems and do analysis ... well, duh some metrics gaming will happen: people avoiding pain.

Probably this will have an opposite affect than what the measurer wanted: people hording things/information/output specifically to improve their metrics, people working around the metrics (outside the managed systems), or essentially spamming the system ("oh you found a bug in my you're reviewing? Can you file a new bug on this?... so my closed bug count for the week goes up...")

[nraynaud]

A long time ago, in western Europe, kids wouldn't go to school if they were sick. To check if they were sick the mother would check their temperature with an alcohol thermometer. Of course the kids understood the goal and used the night lamp to get the appropriate temperature displayed to skip school.

When talking about metric in corporate setting I like to remind everybody that I have a strong suspicion that the guy who invented the thermometer also discreetly invented the night lamp.

(most of the context is disappearing: blinds that force you to use the night lamp in the morning, alcohol thermometers and incandescent light bulbs)

[dsirijus]

Eating raw potato worked too. Eastern Europe, not so long time ago. :)

[JonnieCache]

My memory of this involves dunking the thermometer in the cup of tea.

[watwut]

I did the same. The top of the thermometer exploded as the mercury inside got too large. Apparently, the thermometer was not designed to measure freshly made tea.

[watwut]

Any source of heat, such as radiator, would do the same job.

[mcguire]

And any thermometer. I've used friction on those goofy liquid crystal, flat plastic thermometers.

[nraynaud]

yeah, the trick is that you probably can't get out of bed without being caught.