Combining Data and a Human Touch for Employer Healthcare: A Conversation with Accolade's Raj Sin
Accolade is a technology-supported healthcare company that focusses on helping employees navigate the healthcare system, getting the right care the first time.
By addressing some of the biggest problems facing healthcare – low engagement, lack of trust, and consumer knowledge gaps and errors – they're aiming to provide a drastically better experience and save money for employers along the way.
Raj Singh joined me at the amazing HLTH conference to talk about his work with the growing company.
Accolade argues you can offer more services and simultaneously reduce costs, can you elaborate on that a little bit? Well, people say that 1/3 of costs in the US health system are wasted. That might be unnecessary care, an incorrect care path, or poor compliance with care, but ultimately it’s wasted spend.
In many cases what’s happened over the last five years is that employers have pushed more and more of the load to their employees by giving them a high deductible. The false hypothesis there is that companies are going to take this thing that they couldn’t figure out, and hand it to employees and tell them to figure it out.
Not surprisingly, employees didn’t figure it out and continued to make poor choices. We think that if you can build a relationship with that individual and then leverage technology to extend that relationship, you can get them to make better decisions and when they make better decisions costs come down. It’s a very unique view that says high-touch leads to lower costs and over time leads to a better longer-term relationship.
You’ve been at Accolade now for two and a half years, how do you look back on the growth you’ve had, especially compared to other companies in employer healthcare? What we really focus on is being the single point of entry for the employee: for you or your family trying to understand benefits or trying to understand your health care. We want you not to have to figure out how to navigate but to pick up your phone and make just one call to us, or to pick up your phone and secure message us.
"It’s the data, the recommendations and the human relationship on top. Combining those three together, we have an opportunity to make everything in the system better."
That mind space in the industry up to now has traditionally not been held by carriers — by insurance companies. Our hypothesis was that Insurance companies are good at a lot of things, but that particular element of navigation, advocacy, and high-touch service isn’t really their strong suit. Because of that, we’ve found early success. Now, more and more corporations are grasping that you can do it: you can deliver this type of high-touch service, and your costs will come down.
Before Accolade, you were at Concur, a travel expenses company. What kinds of lessons did you learn at Concur that have helped you at Accolade? I would say the big things you take out of any journey. The Concur journey was 22 years for me, and I think the one thing that was very clear walking in was that Accolade was going to be hard, partly because healthcare is hard but also because business is hard. I learned in 22 years that business has its ups and downs and you have to take a pretty measured approach. Don’t get too high, don’t get too low.
I think the other big thing that I take with me and hold on to at Accolade is that industries can be disrupted, but it takes time. So whether that was a travel marketplace that changed dramatically in 22 years to a healthcare industry that will change slower and be more methodical in its movement, but is equally capable of seeing dramatic improvements in the things that matter.
Accolade seems to stress the use of technology and complimenting that with a human touch. Why did you make that core to your approach? One of the first things we do when an employee uses Accolade is changing the 800 number on the back of the insurance card to ours. So, for starters, they won’t get put on hold because we staff differently and have the different technology to really understand their context.
60-70% of the time someone calls in for the first time, it’s something super tactical. They’re calling because they lost their ID card and need a new one, or they don’t have a primary care doctor and are looking for on their bus route home, or they can’t understand a claim. What used to happen before Accolade would be that people would call in for a new ID card and the plans would treat it like a call center and get them the ID card and get them off the phone in 30 seconds. But Accolade asks a different question: we’ll say ‘Why do you want a new ID card?’
"One of the interesting things we do that’s unique is to give our people the right to overrule the engine. We realize the engine can’t hear that caller’s tone of voice, or the kids screaming in the background."
They’re not getting a new ID card because their wallet’s feeling a little light, they’re getting a new ID card because they’re about to see the doctor. The logical thing we say is ‘Of course I’ll get you a new ID card; what’s going on, are you about to see the doctor?’
The minute you free that human being on our end from just trying to get users off the phone because they have other calls waiting, you can instead try to dig, ask probing questions and understand what the real meat is. The way I like to think about it is that the healthcare industry was taking all these signals about what people wanted in the industry and throwing them away. Instead of ignoring them, we take them and try to learn from them. What’s the role that data plays in customizing care at Accolade? We’ll consume 2-3 years worth of that employee’s claims data. That allows our recommendations engine, our machine learning engine to improve. We can say ‘Here’s what we know about Zach, he spent $100 on healthcare last year, he’s very healthy, he doesn’t have a primary care doc’. So when you call in, our recommendations engine is telling our assistant to just get you your ID card, educate you on your primary care benefits, let you know it’s important to get a primary care doc and get you on his way. Now, that’s using data to help our people really know where to focus.
The next call that comes in may be a 52-year-old who looks like an untreated diabetic and spent $17,000 on healthcare last year. If they call in for a new ID card, our recommendations engine may have ten different question. We want to help this person understand that they need to get a foot exam and get their A1C screened. So the questions our engine will ask this person are totally different.
One of the interesting things we do that’s unique is to give our people the right to overrule the engine. We realize the engine can’t hear that caller’s tone of voice, or the kids screaming in the background. If they overrule the engine all we ask is that they train the machine and tell it why they changed the decision. The machine learns because we watch the choices our health systems and nurses are making and if we don’t understand the choices we ask about them. That allows us to apply those lessons to a cohort that looks like them. —and do you re-ingest the outcomes after decisions? Bingo Accolade can also help direct employees to specialty services like digital therapeutics. How do you use the platform to help decide which services to direct users to? So there are two prongs to that answer. One is that a lot of times our customers have already made those decisions. They’ve chosen Grand Rounds for second opinions, they’ve chosen Teladoc for telemedicine, Castlight for transparency solution, so if they’ve made those choices, we’ll plug them into out platform to help the systems talk openly. That’s something that really doesn’t happen in healthcare enough: open, interoperable data that allows people to share information and provide better care. The industry’s not great at that.
We require it to be an open platform. If you are asking for our recommendations then we’ll put all of the partners through a rigorous checklist that has our CMO review them clinically, we’ll review their financial wellbeing, so we can give our customers safe places to go. Do you have any thoughts on the current status of data integration and where it may go in healthcare? I think the next generation of players is all about data interoperability. I don’t talk to a lot of digital platforms that don’t share data with us and expect us to share data back with them. As long as we get the permission of the consumer and the permission of the employer, the next generation of consumers are into it. I wish that were true for carriers today. I think there are two ways to think about data. 1. Given permission, I’ll share it to provide a more seamless service 2. I have this data, and if I give it to someone else they may take my customer, so I’m not going to share it That second perspective is very last-century and doesn’t yield all the value of the innovation that’s happening within the industry. There are too many places where players aren’t going to give you their data because they’re not innovating, but they want to stop others from innovating as well. That has to change in health, or we’re not going to go as fast as we can to improve our industry. Your clients include plans and employers; how have the conversations differed between the two? Our clients are actually mostly employers, but there are some forward-thinking plans who want to work with us to provide a better experience. In almost every case we drive network adoption up because it’s not a good experience for patients to go out of network and pay $1500 out of pocket.
What we do is good for the insurance company. There are some plans that recognize they may be good at building networks and processing claims but not at the consumer experience and so we’re having more and more positive conversations there. How do you decide where to invest in your product? We start with the consumer first. Sometimes improving the consumer experience means giving our people better data. So we may not always augment the patient app or add new portals but instead improve our machine learning to deliver better insights. We correlate everything to the consumer experience, to NPS, to engagement levels and ultimately to churn. We want our investments to be directly correlated to where we can get the most value. And you just created a new version of the product? We just created a new portal which now has the capacity to tie all of your benefits plans into a single place and leverages our machine learning engine to deliver you the right stuff. If you’re a diabetic, you should see Livongo on the first page. Leveraging our machine learning engine to make our portal very smart and then to make our portal a two-way data engine is the key to improvement. The data and the machine learning engine you have really seem like the bedrock of the patient experience, correct? Bingo. It’s the data, the recommendations and the human relationship on top. Combining those three together, we have an opportunity to make everything in the system better. That’s why we’re seeing success.
How do you work as a company to have a long-term perspective to invest in patients and save down the road? We see cost savings go up the longer we have a relationship with a patient.
—and probably a nice bend there too. Yep! As we get patients moving from unmanaged care to managed and preventative care, we see the trendline continue to bend. It’s obviously a compounding impact because our competition may be seeing a 7% or 10% increase in trend every year. So the core to building that relationship is the human component.
Any interest in Medicare Advantage?
We’re very focused right now on employers. Medicare Advantage is a space we want to play in and will care about over time, just one that comes at a little bit more scale for our company. I understand you got $50 million in funding in March? Who were those backers? Last month actually, April. We had a group of existing investors, Andreessen Horowitz being the biggest, Carrick Capital Partners, and some new investors such as Cross Creek and Madera.
"When Amazon, JPMorgan, and Berkshire Hathaway come into the market and are so tired of the answers handed to them that they’re just going to do it themselves, that’s when you know employers are frustrated."
What do you aim to use the money for? We’ll continue to invest in product and technology. We also want to grow the sales team, grow distribution. We have some customers and want to grow our customer base. We’re serving 1.5 million people today, so we have the capacity to go to our new prospects and say we’ve proven it now. Our handful of customers who have been with us for years can prove our savings. The product has matured as well, so it’s time for us to start elevating our pace a little. Do you have a lives under management goal? I’d be disappointed if I wasn’t sitting here next year and our number wasn’t closer to 2.5 million. Where do you see the employer health space in 5-10 years and where do you see Accolade fitting in there? Employers are going to start taking a far more proactive role. I think the world used to be that you went to your insurance company and bought everything from them, and once every 5 or 6 years you put them out to bid and found a new plan. They would give you a telemedicine solution, a wellness solution, and everything: it’s all rolled up into their platform. The reality is that our customers, our prospects, and the market are smart enough to realize that there’s so much innovation happening that they don’t want to go to one single stop.
They might think a second-opinion solution may be really helpful for orthopedic surgery and they’ll go out and find the best one for them, not just take what’s given to them. Those services can move the needle for them. There are engaged, proactive employers who are going to take the supply chain apart and pick what’s most helpful for them. That’s happening more and more, and it’s why companies like us exist, and why we’re growing. When Amazon, JPMorgan, and Berkshire Hathaway come into the market and are so tired of the answers handed to them that they’re just going to do it themselves, that’s when you know employers are frustrated. We’re going to see that trend continue. It’ll mean the rebirth of things like TPA’s and the rebirth of point solutions and it’ll mean the growth of innovation in that space. If we do our jobs, we should be one of the innovators: one of the open platforms that help employers pick any solution they want and will support them if they want to change solutions and their employees won't have to be re-trained. A lot of recent conversations have happened, like the CVS keynote here at HLTH, talking about how healthcare is best when it’s local when it’s personal. As a more remote solution, how do you feel about their argument? There absolutely is an element of things like on-site clinics, or what CVS aspires to do with Aetna or what Walmart might aspire to do with Humana. The idea of getting to the last mile and engaging that consumer is really sensible.
"We have one of the most confusing and archaic healthcare systems in the world, and there’s a reason for it."
Just because we don’t doesn’t mean it isn’t a good idea. I think we can be supportive of those types of ventures. People have to get out of the business of saying ‘If I don’t do it then it’s bad.’ We have to start saying ‘That’s a good idea and I’ll support it because I want to make people healthier.’ Within your space, what’s a buzzword that you want to see die? Machine learning should probably find a new term; it’s overused What do you think young people aspiring to be in health-tech should learn? I recommend a lot of people joining Accolade who hadn’t worked in healthcare before to know the history. We have one of the most confusing and archaic healthcare systems in the world, and there’s a reason for it. Ezekiel Emmanuel wrote a book called The History of American Health care, and I think it’s a brilliant read. Marty Makary wrote a book called Unaccountable, and I think it’s a great piece about improperly aligned physician incentives. There’s probably 5 or 6 of those books that speak to the history of our healthcare. Our history is both about clinical outcomes and financial outcomes, and those financial outcomes have often outweighed the clinical in our healthcare system. Well, thank you so much, Raj!
A series of interviews with startup founders at HLTH: Check out the interview with OrangeTheory's Dave Long on the future of data-backed fitness
Note: This conversation was taken from a conversation at the HLTH conference in May with slight semantic changes for readability.