Trying to eat better? You’re not alone. You and your fellow travelers spend tens of billions a year on diet-related products and services. It’s a big industry with an extremely questionable track record. Take a step further back to look at health care and we jump a few orders of magnitude more, with annual U.S. expenditures up in the trillions.
It is here that Massive Health, led by CEO Sutha Kamal sees a design opportunity. “Your body is the ultimate interface problem,” says the company’s website. “Sometimes, it just doesn’t give you the feedback you need.”
The company envisions a range of well-designed products that can improve those feedback loops and help improve the daily habits of their future customers. By “well designed” I don’t just mean “nice looking,” though that’s a part of it. At its heart, Massive Health is a big data company and the real design work they are doing is marrying a simple front end to a surprisingly deep back end. Speaking with Kamal, you quickly understand that there is a layered long-term plan being implemented in phases.
To start, they’ve set their sights on your terrible eating habits and the diet industry’s routine failure to fix them. Their first foray is called The Eatery (Massive Health Experiment 01) and it’s an iPhone app that lets you take pictures of everything you eat.
The food diary is a common strategy used by doctors and nutritionists helping their patients eat better. There’s a lot of research that shows that getting someone to record everything they eat drives change, says Kamal. The Eatery is what happens when you take a food diary and connect it to a social network and a massive back end.
Here’s how it works: You snap a photo of your meal and caption it. The app guesses where you are (you can adjust this as well as your portion size) and then you drag the picture onto a sliding scale from “Fit” to “Fat” based on how healthy you think the food is. You are then given the anonymous images of a few other people’s meals to similarly rate. After awhile, the crowd will have rated your meal. Every week, you can go back and track trends with a nice set of visualizations.
If this seems like a kind of wishy-washy way to track your eating habits, Kamal wants you to know that this is on purpose. Most diet apps, he says, try to provide extremely specific information by pulling calorie counts and other data from online sources and previously entered meals. But the burrito you are eating probably has almost nothing in common with the burrito your search pulls up. It’s false precision.
The important thing isn’t the exact number of calories that you are eating, says Kamal, it’s the changes over time. “We don’t care what you had for lunch today. The questions is: Are you eating better this week than you ate last week?” It’s about tracking and encouraging marginal changes and transforming habits.
There is a tradeoff that must be made when gathering data. High-fidelity data like exact portion sizes and other details of your meals might be useful, but the time and effort required to gather that information means that most people won’t do it. The Eatery chooses instead to make it really easy to record your meal, going for high density of data instead. “I get way more meals, even though I know less about each individual meal.”
In fact, says Kamal, the more subjective approach to data gathering, itself, reveals interesting information. This is the moment where the big data mindset of Massive Health becomes clear. Remember the anonymous meals snapped by others that you were rating?
“There’s a lot of other information we know about the pictures we’re showing you,” says Kamal. “While you might feel like all you’re doing is playing a little game, we’re gathering a ton of information about what you think is healthy.”
Some of those images were created by users like you, but others were reviewed by nutritionists. In other words, some of the images you are rating are images for which Massive Health has known good information about the healthiness of the meal. When you rate those images, you aren’t simply helping Massive Health give other users feedback, you are telling Massive Health how good you are at guessing how healthy your food is.
The goal is not only that you will start eating better, but that over time, you will get better at figuring out what’s healthy. By quietly testing you, Massive Health can begin to find out if that’s working. They can see how far you deviate from their expected answer and see if that deviation eventually improves. Some eagle-eyed users have filed bug reports where they notice they’ve been asked to rate the same image more than once. “It’s actually not a bug,” says Kamal. “It’s interesting to see: how consistently are you rating over time?”
As Massive Health continues to run the service and gather users’ data, they begin to be able to do more interesting computation on it. Kamal says that over the next few months, we can expect to see all kinds of new features rolling out that exploit what they’ve learned, to give users nudges in the right direction. “Imagine if you could have a personal trainer who knew you and cared about you who could show up for 30 seconds 10 times a day,” says Kamal. The emphasis is on small regular interventions instead of asking for longer deep engagement from the users.
It’s an intriguing approach, and one reminiscent of how Google, that biggest of big data companies, tamed the web. Rather than trying to create a top-down directory, PageRank found signals in the chaotic noise of the network. Similarly, Massive Health’s approach is to gain insights from the idiosyncratic activity of its users. Both companies also see the value in hiding all of that information behind a very simple interface.
“Everyone talks about big data in health care, and traditionally that means ‘How do I suck information out of electronic health records?’ or ‘How do I get a zillion sensors on your body?'” Kamal says. With this app, Massive Health charts out a different approach. Remember, The Eatery is Massive Health Experiment 01. That number at the end implies there’s a whole lot more to come.
[Top image by MnemosyneM/Shutterstock]