One of my favorite podcasts, a16z, recently created a secondary, news-oriented show, “16 minutes.” 16 Minutes is great because the host, Sonal Chokshi (who also hosts the a16Z podcast), brings on various in-house experts from a16Z’s venture capital firm to provide insight into each week’s news topics. This week, Sonal brought on general partner, Vijay Pande, to discuss the current state of wearable computing. For today’s update, I want to highlight this eight minute conversation (it was one of two topics covered on this week’s episode – fast-forward to 7:45), and build on some of the points Sonal and Vijay make during their chat.
The conversation begins by covering a recent deal struck by the government of Singapore and Fitbit. Singaporeans will be able to register to receive a Fitbit Inspire band for free if they commit to paying $10 a month for a year of the company’s premium coaching service. This is part of Fitbit’s pivot toward a SaaS business, and a stronger focus about informing users about what the data being gathered actually means. Singapore’s Health Promotion Board will therefore have a sizeable portion of its population (Fitbit’s CEO projects 1 million of the 5.6 citizens will sign up), monitoring their data consistently via wearable devices that can be tied to each citizen’s broader medical records.
This then leads to a broader conversation about the ways in which wearables have been maturing, and in many ways, wearables are growing up. To Vijay’s point, we’re moving way beyond step-counting into much more clinically relevant measurable data. Consumer wearables that are increasingly being outfit with more sophisticated, medical-grade sensors, combined with the longitudinal data that can be gathered since they’re being worn all day, creates a combination not yet seen before. Previously, we’ve been limited to sporadic data that’s really only gathered when we’re in the doctor’s office. Now, we’re gathering some of the same types of data by the minute, and at the scale of millions and millions of people.
Other factors: 1) sensor tech greatly improved 2) sensor tech now widely distributed – several 100M units in use just since 2017 3) AI/ML tools now far more accessible 4) healthcare market moving toward value-based care driving more payer interest in preventative care/monitoring
— Ryan Kraudel (@kraudel) September 18, 2019
Ryan Kraudel, VP of Marketing at biometric sensor manufacturer Valencell, made me aware of this podcast episode (thanks, Ryan) and added some really good points on twitter about what he’s been observing these past few years. A big part of what’s different between today’s wearables and the first generation devices is the combination of more mature sensors that are proliferating at scale and the machine learning and AI layer that’s being overlaid on top to assess what the data is telling us, which has become more sophisticated as well.
To Sonal’s point, we’ve been benchmarking our data historically against the collective averages of the population, rather than benchmarking our data against our own personal data, because we haven’t had the ability to gather the personal data in the ways that we can now. When you’re recording longitudinal data over a long period of time, such as multiple years, you start to get really accurate baseline measurements unique to each individual.
This enables a level of personalization that will open the door to preventative health use cases. This has been a big application that I’ve been harping on for a while – the ability to have AI/ML assess your wearable data constantly to help identify risks in your data, based on your own historical cache of data that’s years and years old. Therefore, you enable the ability for the user to be notified of threats to their health data. To Vijay’s point at the end, in the near future, our day-to-day will not be that different but what we’re learning will be radically different, as you’ll be measuring certain metrics multiple times per day, rather than once a year during your check up.
-Thanks for Reading-
Dave
