Mapping the quantified life with wearables

Data science consultancy Profusion ran a research project into the limits and possible application of data collected and analysed from wearable devices. Practice Lead Anthony Mullen explains more.

Anthony Mullen

Guest author Anthony Mullen

What did Profusion do?
For a ten day period, Profusion tracked the biometrics, behaviors and attitudes of 31 employees who volunteered to wear Fitbit Charge HR devices, self-report their emotional state and use tracking software. 171 metrics were tracked for each participant, 24 hours a day, creating one of the most diverse data sets ever gathered using wearable devices.

The purpose of the project was to better understand human behavior, the limits of data tracking on wearable devices, the possible application of the data that was gathered, and fire the starting gun on serious data science research into wearable technology.

What data did Profusion track?
By tracking 171 different metrics, Profusion developed one of the most complete pictures of how much data could be collected and analysed from a wearable device. Information gathered included:

  • Rich demographic details – e.g. age, sex, height, weight, marital/family status and living arrangements
  • Heart rate, activity and sleep – including duration and quality of sleep, calories burned, activity undertaken
  • Emotions, attitudes, stress – participant self-reported their emotional state at two hour intervals
  • Locations, social context
  • Spend across various categories (e.g. electronics, clothing, travel, beauty and health)
  • Food categories – including alcohol and cigarettes consumed
  • Web browsing data – at work and at home, covering mobile, tablet and desktop devices

What did Profusion discover?
At the end of the project the data gathered was gathered and sorted into ten data sets. Using bespoke algorithms, Profusion’s data scientists were able to gain unparalleled insights into the well-being, personality and behavior of its staff.

Some of the key findings:

  • Profusion staff fall into four distinct segments with wearable data proving to be a strong signal in differentiating those segments. We called these segments: Busy and coping, Irritated and unsettled, Life is passion and Accelerator pedals
  • That self-reported stress correlates to heart rate and activity
  • The popular notion about sleep patterns for people with children is wrong
  • Participants had a notably higher average heart rate at locations like pharmacies, supermarkets and clothing shops compared to gyms

The breadth of data collected in this pilot is a gold mine to really working out what makes people tick. This information could be used to power employment wellness programs, provide servicing messages at the right time and in context or power the next generation of segmentation strategies for advertisers.

In search of the quantified self (image: Christopher Michel)

In search of the quantified self (image: Christopher Michel)

What next?
As part of the next iteration of the research project, Profusion plans to:

  • Develop apps to support more effective data capture
  • Include beacons in Profusion’s office space to get richer picture of micro location
  • Send triggering messages to users based on certain conditions (location, heart rate)
  • Include enterprise data
  • Run proof of concept tests for innovative brands looking to explore employee engagement initiatives

About Profusion
Profusion is a data science consultancy specializing in improving how organisations interact with people. Founded in London in 2008, Profusion boasts one of the UK’s largest data science teams. This enables Profusion to analyse large and complex sets of data and to develop bespoke algorithms that answer key business questions.

Profusion has more than 60 staff spread between its offices in London and Dubai and has extensive expertise in identifying consumer preferences, marrying offline and online marketing, attribution modelling, and wearable technology.

/WTW

Leave a Reply

Your email address will not be published. Required fields are marked *

%d bloggers like this: