Where's It Hurt? Healthgrades Examines The Data

There are many big databases in the U.S. But few top that of Healthgrades, the free platform that helps consumers find doctors and hospitals, schedule appointments and communicate with healthcare providers. It pulls in roughly 30 million uniques a month, along with demographic and behavioral data. And it will soon have input from wearable devices.

But it’s no small job pulling this information together to serve hospitals and pharma clients. For input on this, we interviewed Mayur Gupta, SVP in charge of digital at Healthgrades. He previously worked at Kimberly-Clark.  

What’s your business model?
We have a massive platform that engages with more than half the country, where consumers are looking for the right doctor. We leverage data science, both qualitative and quantitative, to drive communication across channels. We’re in a strong position to create lookalike models. And we have risk and propensity models for 200 health conditions. 



How do you get there?
There are two levels. One is healthcare specific; the other is channel-agnostic. First, I want to understand who she is and what her needs are. Today, 80% of healthcare is influenced by non-medical determinants like how much do you sleep and your educational qualifications. We connect all that data with clinical data. We’re agnostic as to what vertical she may be. We can treat her as an email mom, a direct mail mom or a digital mom. She doesn’t see the isolation: She wants the best healthcare at the best prices.

Do you mostly serve women?
It isn’t limited to women, but if you look at the behavioral patterns, we know that Mom makes the healthcare decisions for the family.

What’s your main touchpoint?
Obviously, our is the highest performing channel within our eco system, and the most promising for our hospital partners. It results in Uberization — the patient controls everything. A good portion of traffic comes from search engines, and a reasonable amount comes through our own brand- and direct-to-consumer marketing efforts.

Do you remind patients to take their medication?
We are starting to pilot that. We’re expanding the horizon from finding the right doctor and making appointments to being a companion to take you through the journey. Our mobile app is a key component.

Do you get data from wearable devices?
We are in the middle of some pilots, and having conversations to explore obtaining the computing ability. And we’re working with third-party aggregators that specialize in wearable data.

Don’t you run up against HIPPA and other privacy laws?
It’s at the top of our minds all the time. We know where the boundaries are.

How do you allocate channel performance?
It’s truly contextual. The advancements in technology are enabling us to do that in a new way, while eliminating latency. Before, we had to wait for her to go to the hospital. We now know what she’s reading, what seminars she’s going to. We listen to her, and harvest the signals she’s leaving behind. We also get data from the Epsilons and Axioms of the world. The other aspect is performance analytics. We have redefined attribution models. It’s not an easy problem to solve.  

What are the challenges?
It’s fragmented across the entire eco-system at many brands. Data is stored across multiple systems: It’s beyond most vendors. You plan, measure and optimize by channel. You’ve got the CRM agency, and the social agency. These silos look at the last click — they’re very old school. Some brands are becoming more channel agnostic and horizontal. The data management platforms let us connect the data. The operating models haven’t evolved as quickly.

What can companies do about this?
There are a few things, from planning the experiences in a channel-agnostic way to execution and measurement. True attribution is when you’re taking an attitudinal approach. The channels are influenced by the problems you’re trying to solve. For example, everyone is interested in e-commerce, but that doesn’t mean you need an e-commerce store for an incontinence product: There’s a stigma. The right experience could be a subscription model —you buy it once, and continue to get your supply. Part two is having a platform that enables you to get signals and performance from all these channels.
Next story loading loading..