What Self-Quantification Teaches Us About Digital Analytics
At the intersection of fitness, analytics and social media, a new trend of “self-quantification” is emerging. Devices and Applications like Jawbone UP, Fitbit, Nike Fuel Band, Runkeeper and even Foursquare are making it possible for individuals to collect tremendous detail about their lives: every step, every venue visited, every bite, every snooze. What was niche, or reserved for professional athletes or the medically-monitored, has become mainstream, and is creating a wealth of incredibly personal data.
These aren’t the only areas that technology is creeping in to. You can buy smart phone controls for your home alarm system, or your heating/cooling system. “Smart” fridges are no longer a crazy futuristic concept. Technology is creeping in to every aspect of our lives. This can be wonderful for consumers, and a huge opportunity for marketers, but it has to be done right.
In this series of blog posts, I will explore what this proliferation of tools and data looks like, how it relates to analytics, and what it means for marketing, targeting and the privacy debate.
What Self-Quantification Teaches Us About Digital Analytics
Since April, myself and a surprising number of the digital analytics community have been exploring devices like Jawbone UP and Fitbit. Together with apps and tools like Runkeeper, Withings, My Fitness Pal, Foursquare and IFTTT, I have created a data set that tracks all my movements (including, often, the precise location and route), every workout, every bite and sip I’ve consumed, every minute of sleep, my mood and energy levels, and every venue I’ve visited.
Amidst the explosion of “big data”, this is a curious combination of “big data” (due to the sheer volume created from multiple users tracking these granular details) and “small data” (incredibly detailed, personal data tracking every nuance of our lives.)
Why would one go to all this trouble? Well, “self-quantifiers” are looking to do with their own “small data” exactly what we propose should be done with “big data”: be better informed, and use data to make more educated decisions. Over the past few months, I have found that my personal data use reveals surprisingly applicable learnings for analytics.
Learning 1: Like all data and analytics, this effort is only worthwhile and the data is only valuable if you use it to make better decisions.
Example: My original reason for trying Jawbone UP was for insight into my sleep patterns. Despite getting a reasonable amount of sleep, I struggled to wake up in the morning. A few weeks of UP sleep data told me that my current wakeup time was set right in the middle of a typical “deep sleep” phase. Moving my wakeup time one hour earlier, meant waking in a lighter phase of sleep and made getting up significantly easier. This sleep data wasn’t just “fun to know” – I literally used it to make decisions, with positive results.
Learning 2: Numbers without context are useless.
Using UP, I track my daily movements, using a proxy of “steps.” Every UP user sets a daily “step goal” (by default, 10,000 steps.) Without a goal, 8,978 would just be a number. With a goal, it means something (I am under my goal) and gives me an action to take (move more.)
Learning 3: Good decisions don’t always require perfect data
Steps is used as a proxy for all movement. It’s not a perfect system. After all, it struggles to measure activities like cycling, and doesn’t take into account things like heart rate. (Note though that these devices do typically give you a way to manually input activities like cycling, to take into account a broader spectrum of activity.)
However, while imperfect, this data certainly gives you insight: Have I moved more today than yesterday? How am I tracking towards my goal? Am I slowly increasing how active I am? Did I beat last week? Good decisions don’t always involve perfect data. Sometimes, good directional data and trends provide enough insight to allow you to confidently use the data.
Learning 4: Not all tools are created equal, and it’s important to use the right tool for the job
On top of Jawbone UP, I also heavily use Runkeeper, as well as a Polar heart rate monitor. While UP is great for monitoring my daily activity (walking to the store, taking the stairs instead of the escalator), Runkeeper gives me deeper insight into my running progress. (Is my pace increasing? How many miles did I clock this week? What was my strongest mile?) UP and Runkeeper are different but complementary tools, and each has a purpose. Which data set I use depends on the question at hand.
Learning 5: Integration is key
One of things I enjoy the most about UP is the ability to integrate other solutions. For example, Runkeeper pushes information about my runs to UP, including distance, pace, calorie burn and a map of the route. I have Foursquare integrated via IFTTT (If This Then That) to automatically push gym check-ins to UP. Others have their Withings scale or food journals integrated.
Depending on the question at hand, UP or Runkeeper might have the data I need to answer it. However, there’s huge value for me in having everything integrated into the UP interface, so I can view a summary of all my data in one place. One quick glance at my UP dashboard tells me whether I should rest and watch some TV, or go for an evening walk.
Learning 6: Data isn’t useful in a vacuum
The Jawbone UP data feed is not just about spitting numbers at you. They use customisable visualisations to help you discover patterns and trends in your own data.
For example, is there a correlation between how much deep sleep you get and how many steps you take? Does your active time align with time taken to fall asleep?
While your activity data, or sleep data, or food journal alone can give you great insight, taking a step back, and viewing detailed data as a part of a greater whole, is critical to informed analysis.
The bigger picture
In the end, data analysis is data analysis, no matter the subject of the data. However, where this “self-quantification” trend really shakes things up is in the implications for marketing. In my next post, I will examine what the proliferation of personal data means for targeting, personalisation and the privacy debate.