It seems like everybody's talking about fitness device accuracy these days. A small study at the University of Pennsylvania found that out of clip-on pedometers, wristworn wearables, and apps, the wristworn devices were the least accurate. Another study, from the American Council on Exercise, found a range of accuracy levels for step counting on wearable devices, but found them lacking when it came to tracking calories. And when a BBC reporter recently tried out four popular devices at once, she found their readings didn't line up.
Fitness trackers, though they continue to grow in popularity and add new health tracking features, are apparently not keeping up with their core competency: tracking steps accurately. Unless, of course, that's not their core competency at all.
In their responses to the BBC, Garmin, Misfit, Fitbit and Jawbone all said pretty much the same thing: objective accuracy, in a vacuum, is not the most important thing. They all gave some variation of the same answer: the device is consistent with itself. It gives a valid comparison of your steps yesterday to your steps today, even if that step count is a little bit divorced from reality.
That's a response that makes sense from a company that makes its money selling fitness trackers. But it's also a pretty valid point.
Consumer activity trackers are, at their best, behavior change devices. And the ones that succeed in the market are not the most accurate devices, but the ones that successfully engage consumers and motivate them to change their behavior. We need to look no further than BodyMedia as evidence: one of the only FDA cleared fitness trackers, BodyMedia strived for near-clinical accuracy and ended up exiting the market early through an acquisition by Jawbone, a company that's consistently focused on data and design. BodyMedia's device apps haven't been update in nearly two years. While the devices are still available at some outlets, they are hard to come by.
Exhibit B: Articles like this Bustle post demonstrate pretty clearly what people love about activity trackers. Not everything in that article is flattering for Fitbit and the author is clearly not concerned with the clinical accuracy of her Fitbit's step counts. But the device has done a real number on creating a positive feedback loop that's made her exercise more. And that's the game.
Of course, there is one category of people that should be concerned about a possible lack of accuracy in the activity tracker space, and that's the increasing number of doctors that want to take advantage of HealthKit's integration with both fitness trackers and EHRs to use this data in the clinic.
The Mayo Clinic has famously used Fitbits to track surgical recovery time in its heart patients (though they used the clip-on Fitbit One, which actually scored very high on accuracy in the U of P study). And Beth Israel Deaconess CIO John Halamka has spoken more than once about his goal of using patients' Withings or Jawbone devices to monitor their activity as a proxy for their health.
But even in health use cases like physical rehabilitation or remote monitoring, trend data on a single patient is more important than accuracy across multiple patients. And if doctors really need FDA-vetted, clinically validated accuracy, a whole category of new, clinically-focused devices has emerged in the last few years.
Over at Forbes today, David Shaywitz lauded Apple for scaling down the Apple Watch's lofty aspirations and for "behaving like the adult in the room, and telling all the giddy brogrammers that if you’re going to create a serious product for health, you need to really be sure it’s going to work, and work robustly." And Shaywitz is right, when it comes to devices like glucometers, blood pressure monitors, and blood oxygenation monitors, that people rely on to manage their health.
But when it comes to fitness trackers, the most important thing is still just to get people moving.