This App Uses Cell Phone Data To Track How You’re Feeling

It turns out that the way you use your phone is a huge window into how you’re feeling–emotionally and physically. is using that data to help people track their moods, and help doctors track the health of their patients.

This App Uses Cell Phone Data To Track How You’re Feeling
Medicine via Shutterstock

Cell phone usage patterns say a lot about how you’re feeling. Just ask Anmol Madan, a former MIT Media Lab student who collected 320,000 hours of data from cell phones for his senior thesis and discovered that certain patterns predict the beginning of issues like anxiety and the flu. If you’re depressed, diabetic, or have any sort of chronic illness, cell phone habits can be especially useful; a lack of activity signals to caregivers and doctors that something is up.

advertisement, a startup founded by Madan and MIT alum Karan Singh that we first covered after it won the Sanofi-aventis Data Design Diabetes challenge last year, is leveraging smartphone data to help people with a variety of ailments–including diabetes and heart disease–better manage their moods. And this week, it picked up $6.5 million in a Series A funding round led by Khosla Ventures.

The app runs silently in the background of participants’ smartphones, collecting text message habits, call frequency, and location. All that data is analyzed and sent back to both patients via the app and doctors and researchers via an online dashboard. If you suddenly stop calling your friends, or don’t go to work for a few days, that could be a sign to your doctors that they need to check in on you more aggressively.

Doctors also have the option of sending out simple surveys–i.e. how did you sleep last night on a scale of 1 to 10?–using the app daily, weekly, or using any other time interval. (Privacy advocates can rest easy, the app doesn’t track who is being called or texted, just that the calls and texts are taking place. It doesn’t track exact location, either, but it can guess whether a place is work or home based on the time of day and length of stay.)

During a visit to’s San Francisco office, user experience lead Sabih Mir walked me through the participant and caregiver views of the platform. On the caregiver side, “we wanted to try to keep it organized around patient activity and getting a general understanding of overall statistics,” he explains. “It could be researchers or doctors managing a population of patients.” That means it needs to be easy for caregivers to quickly scan through statistics.

The dashboard looks broadly at patient interactions (calls, text messages) and locations, with the option to dive deeper into individual patient activity, including whether they have completed all their surveys. Potential issues–like a patient failing to complete the last three surveys sent–are flagged, but it’s up to caregivers to follow up. There are no judgments on activity abnormality, either. “We try not to dive into behavior patterns. We show the patterns but try not to say ‘this is normal or abnormal.’ It would be up to a researcher to go to [a participant’s page],” says Mir.

For the patient-centered mobile app,’s has two main goals: data transparency and insight into the larger patient community. Patients have access to all sorts of data points–the app might inform a user that they sent and received three texts yesterday, but average 10 texts a day. It can also make broad assumptions, like whether a patient has a good work-life balance (based on how long they spend each day at the location that the app infers is a work location).

advertisement is attracting significant attention from the medical community. “When we won the Data Design Diabetes competition, we were still unclear whether this was for providers, [pharmaceutical companies], or someone else in the ecosystem as an enterprise customer,” says Madan. “We realized the real market is anyone who is managing a population–most often a care provider-type institution.” So, for example, an institution working with 10,000 diabetic patients is who the app is targeting.

The startup is about to kick off a pilot with Sanofi-Aventis for diabetes patients in the Bay Area; a health care system in North Carolina (Novant Health) is launching a pilot in December for a couple hundred diabetes patients, set to scale up next year. In the first quarter of next year (most likely), Novant will expand the program to heart disease patients. is also working on chronic pain management with other providers.

Singh and Madan stress that the platform isn’t just for well-off patients, even though it only works on smartphones. As Madan points out, today’s smartphones can be relatively cheap, and in many cases, they’re the first electronic devices that people choose to buy. Half of all adult Americans own smartphones or tablets. Still, Americans between 45 and 54 that make over $100,000 annually are 3.3 times more likely than people who make under $15,000 per year to have a smartphone. has no shortage of interested parties. Says Madan: “There’s a gap between the number of leads we have that are qualified and the number we can support.” The app is free for patients; institutions are charged a fee per user per month.


About the author

Ariel Schwartz is a Senior Editor at Co.Exist. She has contributed to SF Weekly, Popular Science, Inhabitat, Greenbiz, NBC Bay Area, GOOD Magazine and more.