CASE: BIOE273-3
DATE: 09/01/2017
USER-FOCUSED IDEATION AND DESIGN: [Link]
“To really move the needle and change the way that healthcare is delivered for behavioral health and
increased access, the human component and human touch is critical.” 1
– Omar Dawood, Medical Director, [Link]
[Link] was founded by MIT researchers Anmol Madan and Karan Singh to address the toll of stress,
anxiety, and depression on people’s everyday lives. The pair met at MIT’s Media Lab, where Madan’s
research focused on tracking data from students’ mobile devices to identify illness. Together, they
envisioned better ways to use such approaches to treat mental health conditions. Ultimately, by staying
focused on the needs of their users, they succeeded in making mental health care more effective,
accessible, and affordable to hundreds of thousands of users.
Background
Before he co-founded [Link], Anmol Madan was a researcher at the MIT Media Lab. There, he
collaborated with Dr. Alexander Pentland and researchers at MIT’s Human Dynamics Lab to study reality
mining; the collection and analysis of human behavior data generated by monitoring technology usage.
Specifically, by examining the passive digital footprint created by mobile phone usage, the researchers
could identify predictable patterns and draw conclusions about what people do, where they go, and with
how much they communicate.
When this research also yielded findings about how people with depression behave when they are
symptomatic, Madan became intrigued with the idea of applying these techniques to a wide range of
disease areas from diabetes and mental health. The premise was that behavioral data gleaned from
smartphones, such as movement, location, and call and text frequency and duration, could be used to
establish a baseline, or normal behavioral pattern for an individual. Significant deviations from that
baseline could, in turn, indicate the onset of symptoms. “If someone is depressed, for instance, they
isolate themselves, have a hard time getting up to go to school or work, they’re lethargic and don’t like
communicating with others the way they typically do,” Madan explained. “Turns out, you see those same
features change in their mobile phone sensor data in their movement, [usage], and interactions with
others.”2
By passively monitoring this flow of information, Madan believed it would be possible to detect when
patients were experiencing warning signs and symptoms of an underlying condition and then trigger an
alert to help patients or providers take preventative measures or begin prompt treatment. In this way, “big
data” analytics could be used to move care delivery towards earlier, less time and resource-intensive
interventions. “Acting on that information while it’s still the right time, when patients can correct their
behavior, could certain change the way we deliver care,” Madan said. 3 He recruited Karan Singh, also at
MIT, as a co-founder, and the pair began researching needs that could potentially benefit from this
capability.
This case study was prepared by Samantha Madala, Stacey McCutcheon, Lyn Denend, and Marta Gaia Zanchi. Thank you to
Jeremy Johnson and Omar Dawood for their contribution to this project.
Researching and Validating the Need
After extensive research, the founders ultimately gravitated toward needs in the field of behavioral and
mental health. For one thing, the mental health market was huge. Mental illness affects millions of
individuals in the US every year. Annually, an estimated 43.8 million adults, about 1 in 5, experience
some sort of mental health issue.4 However, access to mental health treatment is limited and patients
remain significantly underserved. Despite the pervasive incidence of mental illness and the massive costs
associated with treatment, 56 percent of adults suffering from mental health issues receive insufficient
care, or no care at all.5 Barriers includes a significant shortage of mental healthcare providers, 6 as well as
cost, distance to the nearest clinic, or other similar obstacles.
Coming Up with the Right Solution
Although the need was clear, Madan and Singh’s vision for an effective solution was initially less certain.
However, based on what they had learned, they were able to come up with a handful of critical need
criteria that would ultimately help them guide ideation and solution design.
The first must-have criterion was directly related to the population they were focused on serving; people
with some underlying mental health concern who wanted help managing their wellness over time. 7 The
founder’s believed that individuals in this population were not especially attentive to the physical or
emotional warning signs that could indicate a mental health issue, so they would benefit from a “human
check engine light”8 to alert them to key signals. However, because most people aren’t particularly
motivated to make major behavior changes to monitor their wellness,9 any solution would have to be
unobtrusive and easily integrated into their normal patterns and routines.
Another important criterion was ensuring privacy for the individual seeking help. “Talking about
depression remains a challenge,” said Madan.10 In speaking with physicians, therapists, and especially
patients, the founders observed a perceived stigma around mental health that made some individuals
reticent to raise concerns or share their symptoms. The [Link] solution needed to be discreet, secure,
and confidential.
Similarly, the founding team recognized that their solution had to make care available around the clock,
whenever the user needed it. “There’s a complete mismatch right now in terms of care and resources,”
said Omar Dawood, who became [Link]’s medical director. “Symptoms are completely unscheduled
whereas care is scheduled. And those two don’t go well together.”
Finally, Madan and Singh wanted to design a solution that would make care both more affordable and
more broadly accessible. This is where their strong technical expertise would be essential. “We cannot
train enough psychiatrists or hire enough therapists to ensure access to high-quality care without
technology playing an important role,” Madan said.11
With these parameters to guide them, the founders undertook a human centered ideation and design
process that ultimately led to the creation of [Link]’s first consumer product. By focusing on a smart
phone app they found they were able not only to take advantage of Madan’s original research but also
address their other must-have need criteria. After downloading the app, patients would start by filling out
a short questionnaire to provide information about their conditions, past and current treatment, medication
use, and healthcare provider(s). The app would then begin passively collecting millions of data points
around cell phone usage.
2
After recording several days of typical use patterns, the algorithms looked for significant deviations and
then responded by sending a text to the user and/or the specified family member, friend, or healthcare
provider to alert them to a potential problem. “When they get the alert, they call and see what’s really
going on. Then, they make that decision of having the patient come in to see a doctor or handle the issue
over the phone,” Dawood explained.
He continued, “The initial concept was to develop the world’s best behavioral health analytics engine,
leveraging it with the most ubiquitous sensor that we have around us, which is the smart phone. There
are so many different, isolated actions that one takes with a smart phone that are not content related...
when you combine them and then match them up against billions of other data forms collected from the
population, you can make patient-specific insights.”
According to Jeremy Johnson, [Link]’s head of engineering, delivering their service through an app
created some additional user-centered design challenges for the team as they conceptualized their
solution. For instance, he noted, they knew they would be competing for user attention against “...every
other mobile app that a user is likely to [open],” including social media sites. To address this, the team
worked hard to design a technology that ran effortlessly in the background, rather than requiring frequent
input from the user. The passive nature of their tracking algorithm was a benefit in this regard, and they
were able to build upon this by automating alerts and other important features.
Another important design consideration, one that all digital health companies must contend with, was
whether or not the [Link] team was developing a solution that would be considered a medical device
by the US Food and Drug Administration (FDA). Apps that claim to prevent, diagnose, or treat specific
diseases or conditions are generally considered medical devices and are, therefore, subject to regulation
by the FDA. In contrast, those that focus on gathering and assessing data to provide guidance or
coaching (rather than diagnosis or treatment) may be exempt from regulatory requirements.
To help them choose the appropriate path, the [Link] team carefully evaluated the core functionality
that they were most passionate about providing to their target users; alerting them to the potential need
for care. Given this focus, they determined that they would not meet the FDA’s guidelines for
classification as a medical device. While this decision eliminated certain regulatory hurdles, it
simultaneously imposed design constraints on the team to ensure that the app would not overextend in
any way to diagnosis or treatment. As Dawood put it, “We needed to stay outside the bounds of being a
medical device, so the product and our messaging is explicitly designed that way.”
Need … and Solution … Evolution
Roughly four years after initial launch, [Link] had established a strong user base and was making
meaningful strides towards its mission. However, through feedback from its users, the team recognized
that they could be doing more to improve mental health management for their community. “Technology in
itself cannot change medicine,” said Dawood. “We had to go back to something that we have centuries of
data on in behavioral health, which is the human touch.” For instance, many users already had mental
health providers they could engage when they became symptomatic, but others did not and would benefit
from assistance in finding someone appropriate to help. As Madan described, “In our drive to improve
care and accessibility for patients with depression and anxiety, we…recognize the fundamental
importance and value of trained mental-health professionals…from primary-care providers and
psychiatrists prescribing medications, to therapists and behavioral coaches…. The challenge we faced
was how to include human care in a scalable manner.”12
3
Ultimately, the team implemented a collaborative care model, where every user is matched with a
personal health coach. “Having a trusted relationship at the core of what we do has become critical,”
Dawood said. The coach serves as a central source of information and a personalized care planner,
linking patients to experienced mental health clinicians such as licensed therapists and medical providers
as needed. “[The health coach] is able to connect a patient to the right sort of health-care provider at the
right time, thus playing a pivotal role in balancing patient needs with available resources,” described
Madan.13 Concurrently, in-application activities based on clinically-validated strategies such as cognitive-
behavioral therapy and mindfulness help individuals better manage their own mental health.
According to Dawood, the addition of human coaches to the offering represents the natural evolution of
the product, while also helping better fulfill the team’s original must-have criteria. For example, care is
available exactly when it is needed. “The coach can elevate people to different levels of care, from
licensed therapy to board-certified psychiatry, exactly when the user needs it. Not two weeks from now,
not tomorrow, but right now, exactly when your symptoms demand it,” he said.
To date, [Link] has been able to reach hundreds of thousands of users seeking to more effectively
manage their mental health. The company has also raised more than $28 million in funding to further their
mission of providing real help to real people.14 Importantly, [Link]’s combination of technology and
human-centered care has elevated the effectiveness mental health management by eliminating a chief
problem in the field: the reliance on patient-reported outcomes (PROs). Explained [Link] seed-round
funder Jon Callaghan, “PROs [have been] the standard collection method for obtaining information about
patient mental, physical, and social health status, and are used to assess risk, aid in clinical decision-
making, and serve as endpoints in clinical research for the discovery of new therapies. Despite their
importance, PROs are burdensome, incomplete, and biased, compromising clinical research and care
decisions.”15 In other words, healthcare insights drawn from objective data are more reliable and valuable.
“With our data collection, it’s like having the psychiatrist sitting in the patient’s living room at all times
being able to collect contextual information,” said Dawood.
Key Insights
• Even when research/technology gives you an idea, the need still matters!
“Don’t start with a solution. Start with a problem and start by understanding why that problem is a
problem. Then you can start looking at plugging in how mobile or other technologies can fit into
that. Too many times, entrepreneurs run with a solution before they have truly understood what
the problem is.”
–Omar Dawood
• Winning solutions are optimized for a target user
When developing a mobile app, the key is to optimize it around what the user needs, not just
what the technology can do. Establish real empathy for your users and design a solution with
functionality and features that will delight them.
• Users and their needs will evolve…and your solution must change with them
Just as [Link] shifted from a mobile monitoring platform to an integrated mental health
approach, digital health companies should be ready to adapt to keep pace with the evolving
needs of their users.
4
1
All quotations are from interviews conducted by the authors unless otherwise cited.
2
Rob Matheson, “Mental Health Monitoring Goes Mobile,” MIT News, July 16, 2014,
[Link] (August 9, 2017).
3
Rob Matheson, “Mental Health Monitoring Goes Mobile,” MIT News, July 16, 2014,
[Link] (August 9, 2017).
4
“Mental Health By the Numbers,” National Alliance on Mental Illness, [Link]
Health-By-the-Numbers (August 28, 2017).
5
“Mental Health By the Numbers,” National Alliance on Mental Illness, [Link]
Health-By-the-Numbers (August 28, 2017).
6
Lousie Radnofsky, “Where Are the Mental Health Providers?” The Wall Street Journal, February 16, 2015,
[Link] (August 28, 2017).
7
Brian Edwards, “Why [Link] is the Most Innovative Digital Health Startup,” [Link], July 18, 2012,
[Link] (August 26, 2017).
8
Brian Edwards, “Why [Link] is the Most Innovative Digital Health Startup,” [Link], July 18, 2012,
[Link] (August 26, 2017).
9
Brian Edwards, “Why [Link] is the Most Innovative Digital Health Startup,” [Link], July 18, 2012,
[Link] (August 26, 2017).
10
Anmol Madan, “Your Smartphone, Your Therapist?” Huffington Post Blog,
[Link] (August 24,
2017).
11
Anmol Madan, “Your Smartphone, Your Therapist?” Huffington Post Blog,
[Link] (August 24,
2017).
12
Anmol Madan, “Your Smartphone, Your Therapist?” Huffington Post Blog,
[Link] (August 24,
2017).
13
Anmol Madan, “Your Smartphone, Your Therapist?” Huffington Post Blog,
[Link] (August 24,
2017.
14
Sarah Buhr, “[Link] Adds $20 Million in Funding,” Tech Crunch December 23, 2014,
[Link] (September 1, 2017).
15
Jon Callaghan, “Disrupting Health Care with [Link],” True [Link], November 28, 2012,
[Link] (August 31, 2017).
