cal_count

The Story of You

At Jawbone, we search for understanding in every piece of data, and knowledge in every number. That expertise helps you make smarter choices.

It seems straightforward, but making sense of the numbers is rarely easy. To do it, a team of data science wizards joins forces with artful storytellers. Sometimes we aggregate UP data from around the world. We’ve examined changing sleep habits during the World Cup, and commonly paired foods. We’ve even realized which country takes the most steps.

But Jawbone doesn’t just aggregate data, we often personalize it in UP’s Insights. Here’s an example:

 


Screen Shot 2014-10-02 at 4.03.08 PM

UP also sends Insight Reports, an incisive story that aggregates one month of data and exposes hidden trends. March 2014’s Insight Report studied the relationship between an individual’s bedtime and the amount of steps they took the following day.

To create this Insight Report, the team started with a hypothesis. “Yes! There is a relationship between bedtime and next day steps. In fact, it’s an inverse correlation!” We hypothesized that people with earlier bedtimes would take more steps the next day, reasoning that an earlier bedtime would lead to more energy. More energy, more activity. More activity, more steps.

An inverse correlation did exist for some people. For those individuals, going to bed one hour earlier corresponded to about 130 more steps the next day.

But that was only the beginning. We found positive correlations. The later those people stayed awake at night, the more steps they took the next day. Others exhibited a very weak correlation that was neither positive nor negative. Some had almost no variability in steps or sleep at all.

We took that data into consideration and sent an Insight Report in UP that was relevant and insightful. One version of the finished product (sent to people with an inverse correlation) looked like this:


Screen Shot 2014-10-02 at 4.15.27 PM

And yet, there’s another catch. The Insight Reports revealed correlations between bedtime and steps, but we’re still not sure about the causality. We don’t know whether someone steps less because they went to bed late, or if that person is going to bed later because they don’t have to take many steps the next day.

We chase causation, and we think we’re close. In addition to Insight Reports, UP also sends challenges called Today I Will. While Insight Reports reveal habits, Today I Will uses a person’s pattern to encourage healthy sleep, move or eating changes. In one example, Today I Will challenges a person to take more steps because their recent step average has fallen from their monthly norm:


Screen Shot 2014-10-02 at 4.17.01 PM

This sound, timely inspiration can prompt smart choices. Users who commit to a step Today I Will are 35% more likely to hit their step goal. They also take 332 more steps than those who do not receive a Today I Will.

Is timely encouragement the most responsible way to inspire healthy behavior? We continue the search because we want to tell your story; and no story is simple.

Special thanks to Brian Wilt (@brianwilt) for help with this post.

 

About The Author

Jessica Caimi

Jessica breathes, moves, eats and writes. She is also particular about comma usage.