As last year was ending, I set some over-ambitious goals for my personal objective of Becoming Faster. The goals, or Key Results in OKR-speak, were simply a new set of personal records for my favourite race distances; 5k, 10k, and a half-marathon. Highly measureable.

The hypothesis for 2017 is that if I run more, I should become faster. Very contrarian, I know. At the end of last year, I had clocked a total of 755 kilometers. This year, the aim is at least 1000 kilometers, about 20 km per week. In contrast to the Key Results, this goal is highly achievable and not a stretch goal. (At Spotify, we’re calling the bits and pieces that you believe will take you toward a Key Result a milestone.)

Is tracking distance a vanity metric, as opposed to an actionable metric? Yes, if it’s treated as a measurement of success towards the objective of becoming faster. However, if it’s treated as a parameter in an A/B test, where 2016 was a ‘control group’ and the 2017 was the test, we’d be able to confirm the hypothesis in events occurring during 2018. Perhaps, we could confirm it even earlier. Basically, The expectation is that the increased distance per week would lead to better results.

There’s obviously an array of variables that’d need to stay static for this experiment to be even slightly scientific - like age, weight, muscle composition, amount of interval training, and probably a number of others. Let’s assume they’re constant.

My Strava