We know why people buy consumer wearable devices – they are trying to improve their health or fitness, track an illness, change their habits, and so on. And, of course, many wearable devices have failed to meet these standards completely.
However, the market failures of the past have actually given us a false sense of security - many of them failed for cosmetic or commercial reasons (remember the Neptune Pine?), which gives the impression that all will be solved when the right slick-looking, well-supported hardware/software combination comes along.
The question that is more fundamental than “are they stylish?“ or “will people buy them?“ is “how do they work?”
“Work” in this sense requires a device to make two very difficult transitions, far more difficult than having Jean-Paul Gaultier design the strap.
First one: the transition from data to information
Second one: the transition from information to action
One at a time now.
Lots of information exists in the intact body that we can access, most of it from electrical signals derived from the brain, the heart, the skin, the stomach, and even the eyes.
Advances in miniaturization and data access mean that it’s more available than ever. But, assuming that we have access to it, what does it mean?
For instance, many devices will give you an electrocardiogram, which we can turn into heart rate intervals. But we do this primarily because we are interested in the state of the autonomic nervous system which determines the heart rate.
Is that transition simple? No. It’s actually full of unresolved scientific debates. The primary metrics of heart rate variability, which so many wearables measure, are the product of a process which we can’t explain.
It is a fatal mistake in any measurement task to assume that just because you have numbers organized, you are entitled to have them mean something.
Even if we have bodily signals we can trust, and we can turn those signals into information, we are still faced with the task of turning that information into usable information.
To turn information into action, signals from wearable devices need to be turned into advice, guidance, behavioral suggestions, ‘nudges’, etc. – it is not simply enough to track the numbers themselves without context. What will you do with the knowledge that you have a waking heart rate of 85 beats per minute? What action does that compel?
Often, not very much action at all – somewhere from a third to a half of all fitness tracking wearables are either quickly abandoned or never used at all.
In many builds, behavioral modification following measurement seems like an afterthought and an eventual job which is more for behavioral scientists than engineers.
However, none of the above is a reason for despair. Multiple elegant builds which successfully record the signals needed to answer the above exist now and are being improved constantly.
In short, we are finally in a position where we have a realistic chance of addressing these challenges. The next 2 to 5 years should be interesting as interest progresses from beyond build quality and towards the meaning and deployment of the signals those builds record.