At a recent evening meetup a friend was describing to me three types of data analytics:
- Descriptive – What is happening now?
- Predictive – What will happen?
- Prescriptive – What should I do about it?
As you might guess analytics in this sense is a continuum. Start off with descriptive metrics and increase your analytics maturity by moving up the scale to predictive and eventually prescriptive numbers.
After the meetup I came home, a bit later than the usual weeknight. So as I walked into the house I was greeted by blast of stuffy hot Georgia summer air.
The Nest thermostat’s auto away had beat me again!
And that’s what makes Nest so much different than other thermostats.
It’s not the amazing industrial design. Yes, it’s beautiful relative to its counterparts.
Nor is it the ability to set temperature schedules from your phone or the web. Although I’m glad I’ll never again have to program an “old school” thermostat. Stick a pickaxe through my foot instead please.
The traditional thermostat is only descriptive. It tells you what the temperature is and what temp it’s set to.
The Nest, on the other hand is not only predictive, but also more importantly prescriptive. It is a behavior changing technology.
Why does this matter?
When considering wearables and IoT devices, the most successful ones will move up the analytics spectrum and become more than just descriptive devices. They’ll be smart.
Think about the traits of wearables:
- They’re always on,
- They’re always with you (or where you want them to be), and
- They’ve got a range of built in sensors
…so as a result they can collect tons of data.
But having data and making it meaningful are two very different things.
Otherwise you’re just left with a bunch of vanity metrics.
The Android Wear combined with Google Now comes to mind as another great example.
- Descriptive – What time is it?
- Predictive – Knowing you have a flight today so reminding what time it’s at.
- Prescriptive – An alert on your wrist telling you traffic is bad and you need to ASAP to catch your flight.
Of course this analytics spectrum plays out in many other technologies and industries. In wearables and IoT I’ve so far seen it play an especially important part. Going back to my previous post about how Wearables Aren’t Quite There Yet, building prescriptive analytics into wearables might tip the needle enough that utility and function outweigh many of the aesthetic and technology challenges.