Preventative healthcare and fitness data

Company incentives for growth and the modularity and openness of fitness data are in conflict.

This post tries to argue that if fitness trackers become even more ubiquitous and we adopt them as early as possible in someone's lifetime, the data has a compounding value in deriving insights, but those insights come with the promise that you'll stick around with the company for a long time.

Preventative healthcare uses 1) in the moment adjustments to your environment and 2) spotting trends towards negative outcomes ahead of time.

Let's see a couple of questions that you can possibly ask, on different levels of data:

Example 1: Tracking calories daily, weighing yourself monthly, for a year

Questions you could ask:

  • What did I weight on that day, in that month?
  • Did I do anything during last summer that made me loose that weight?

Questions you can't get an adequate answer for:

  • How much weight did I loose compared to last month? (Weighting once per month gives you a snapshot of that day, but doesn't account for your day to day environment. Loosing 5 pounds can easily be overwritten in your monthly weight-in by things such as stress, high-carb intake, lack of sleep, hitting the gym hard or not.)
  • What does my weight usually do during holidays?

Example 2: Tracking calories daily, weighting yourself daily, for 2 years

Questions you could ask:

  • Am I loosing to gaining weight, on average, based on this number of calories? (You can account for seasonality by keeping the outside factors as steadily as possible - such as eating in your calorie range for 4 weeks - and then averaging the weight loss for that month and seeing where does the weight trend.)
  • What happens to my weight the first week after Christmas? (You could get over the "I gained 5 pounds after the Christmas day" by noticing your daily weight-ins spike and then go down once you return to your normal calorie range and do an average for the 2 years.)

Add things such as monthly blood pressure tracking, activity data and sleep data, do it for 5 years and you could run some pretty interesting analysis on that, given that the data is as accurate as possible. You could open the floor for questions such as:

  • Does my blood pressure increase as I age?
  • Does my blood pressure decrease as I loose weight or increase activity level?

The more data we have, the better questions we can ask, the more we can prevent classic things such as high blood pressure sneaking in on us. Or by quantifying the decreasing level of activity, we might get off the slippery slope of becoming a couch potato sooner.

Looping back to the data ownership: I have an Apple watch, for 5 years, and I track my calories in MyFitnessPal, and I loop them all through some other aggregator. Happy days. And I want to quit. Can I quit? For sure. Do I have access to the data? For sure. Is it easy to transfer it somewhere else? Doesn't seem like it.

MyFitnessPal requires a premium account to download your reports
Screenshot of the FitBit data export file structure. How do you move your data somewhere else?

A problem with preventative healthcare: if it isn't easy and almost ubiquitous, it's hard to incentivise people to do. You don't have enough fire burning under your ass to care about it. So things like raw CSV exports and SQL queries are left for the minority of the early adopters.

What are the incentives for a company to have their data easily port to another provider of the same service? Not that many.

With users and data you can:

  • Make user retention easier, using data moats to raise the switching costs and therefore, needing to keep less in check with our competitors
  • Monetise with better targeted product placements / advertisements based sorting the user base via demographics
  • Train machine learning models
  • Give access to aggregated data to other companies that find direct value in it

Monopoly also comes into play. Consumer electronics is a pricey field to play in.

As a user, when you pick one of these apps, you're not only saying "I like that you make your watches sleek.", you are also silently saying "I decide to trust that you are going to make the right decisions not only today, but also for the foreseeable future, with both your product and my data."

But the big companies are carrying multiple battles:

  • Maintaining current customer and revenue base to keep the lights on (current employees, production costs)
  • Trying to predict who is going to f*ck them over (R&D, new product market fit initiatives)
  • Trying to predict what is going to f*ck them over (sudden changes in the barriers to entry: see Monzo and the well established banks)

I love the current players in the market, don't get me wrong. They are pushing forward fitness and health tracking, and I doubt there is a hidden group of cloaked board members sitting in a room and mapping out how to sell any data.

I rather think this trend is a result of company incentives misaligned with the longer-term benefits of some of the amazing things they are building now. But what the short history of tech has told us so far is that even the biggest players can get sucker punched.

In the game of health tracking and your data, a company shutting down turns consumers into collateral damage.

Imagine being able to do  this,  but across all the possible metrics we can track. The solution is not obvious, and I'm not accounting for disruptive technologies coming in and changing the way we interact with technologies.

What could be some of the qualities of "new data"?

  • data ownership closer to the person who creates rather than the entity who aggregates
  • has the ability to be well understood by non-technical entities
  • modular and plug and play with different health services

I'm ranting out at this point. Maybe I'm describing a data storage standard that companies can link into, but never fully own. Maybe I'm thinking about a decentralised health tracker that plays in the consumer space.

Subscribe to tibi iorga

Don’t miss out on the latest issues. Sign up now to get access to the library of members-only issues.
jamie@example.com
Subscribe