Some thoughts on Average Pace, and the usefulness of data
I love data. Ok, maybe not as much as Hilary Mason, but I'm still a big fan. I recently started using a GPS watch while running, which has added a new column of data to my personal collection. I had an important learning experience today: Average Pace is the most important metric to monitor while I’m running.
While the watch will attempt to spit out an Instantaneous Pace measurement, it's useless. Anything that updates faster than the Average Pace has too much fluctuation to be useful. I can't sort out the signal from the noise. Anything that updates more slowly (such as Lap Pace) is updated too infrequently (once every 8-minutes, if I'm hitting my target pace) for me to make adjustments. (For the curious, it's possible that Average Lap Pace would fall into the magic range of usefulness, too. I'll try it on my next run.)
This lesson translates over to my personal dashboard as well. In order to be useful as a decision-making tool, it needs to update at a particular rate.
- What I really need to know for food: am I where I am supposed to be for this time of the day? Kind of like a speedometer, I want to keep my pace of eating in a particular range throughout the day.
- Same goes for sleep. I don’t really care about my lifetime average, or how much I slept last night. What really is actionable is how well I’m doing for the week. It’s the happy, meaningful middle.
I think there's a good likelihood that this same swath of data usefulness will apply to tracking (and maintaining) my performance at work. If I need to make a certain number of sales/business development contacts per day, what is the reasonable data rate that I need to adjust my pace? Every 2 hours? Right now, our business team lives on a daily schedule. We have a list of activities, and at the end of the day, we give ourselves a percentage score (number complete divided by number scheduled). What about for our development team? They are currently following the Scrum method of Agile development, working in 3-week sprints. This method specifies a 15-minute Standup meeting every day. Is this barely frequent enough, or is it distracting?
What I'm getting at is that there is a limit to the usefulness of data. There is an optimum range of actionability that varies for different people and for different tasks. I'm looking forward to doing some experiments to find out just where that is.