Episode #109: Tame Your Martech Frankenstack with Lessons from Healthcare
Success in CXM requires people, process, and technology. You need people who embrace a customer-first mindset, and a culture that encourages it. You need playbooks that encapsulate your CXM process. And you need technology to make it all work. That’s where we can look to the healthcare industry. They’ve gone through a best-of-breed vs. best-of-suite transformation, and we can learn a lot from their experience.
See all Copernican Shift podcasts
The Copernican Shift on Apple Podcasts
All right, welcome to the CXM Experience. Today’s analogy day. This is going to be super fun. I’m looking forward to this show. I’ve wanted to do this for a while. And it’s something I actually talk about a fair amount when I’m doing really deep background on myself with a customer. So, I’m going to share it with you today. As always, I am your host Grad Conn, and I am the chief experience officer, CXO, at Sprinklr. I’ve been at Sprinklr for a while now. I was Sprinklr’s first customer, signed my first contract for Sprinklr for $30,000 in February of 2012. So just a shy over nine years with the Sprinklr team in one form or another and I’ve been at Sprinklr — in a week — I’ll have been at Sprinklr for three full years as an employee. So as either a customer or employee nine years, three years at the helm, and it’s been quite a journey.
I want to talk a little bit about my journey over the last, I guess 15 years. It really relates a little bit to how you think about architecting experience from a tool standpoint. Now, I’m going to delve a little bit into the definition of CXM for a second. I want to put some definitional stakes in the ground. We’ve done this before on the show, but not for quite a while. We’ve been blessed with a lot of guests recently. It’s been great, and lot of them are touching on these topics.
First of all, I want to note that CXM, or customer experience management, is not viewed as a category by Gartner and Forrester and Constellation and other analyst firms. It’s not viewed as a category. It’s viewed as a process, which is sort of interesting. Now, I don’t know if that’s true or correct, or will be true long term. That’s where we are today. Now the reason that they say that, is that if they think about CXM, CXM is broader than any tech stack. There’s a people issue. There’s a process issue. And there’s a technology issue.
For example, you could put in an amazing CXM platform, like Sprinklr, the world’s greatest, most amazing just awesome CXM platform. And if you haven’t trained your people, and you don’t have any processes in place to use it correctly, it’s not going to amount to a hill of beans. I’ve been watching a lot of westerns lately. So “hill of beans” just sort of hit me. Sergio Leone is my current director of choice. I’m currently watching “Once Upon a Time in America.” If you’ve ever seen this movie, it’s crazy. Definitely worth a watch. Although it’s four hours. It’s got an intermission, but it’s a bit of a long one.
Anyway, coming back to the topic. So CXM, people, process, and technology. Think about Disney. There was no CXM platform when Walt Disney opened Disneyland in 1955. But boy, they were focused on experience. Everyone picks up the garbage. Everyone’s focused on the guest experience. The teammates, they’re called cast members. Everyone’s on stage all the time. I had an opportunity once to spend nearly a year in Disneyland every day — that’s a story for another time — and just sat and watched and watched. And you’d see obviously very senior people wearing suits and ties, walking along, saw a piece of garbage on the ground, would lean over, pick it up, put it in a receptacle nearby. There was never like, hey, call over the person who does the garbage, or call in this issue, or circle it. See the problem, deal with the problem. And you’d also see lost children and angry adults and all sorts of people with issues approaching any Disney employee and immediately being looked after and never being sent off. That’s a massive hiring and people issue. That is training, that is culture, a massive part of that is culture. That’s just the way we do things.
Then there’s process. You’ve got to have great playbooks, you’ve got to be able to have things that can be instantiated and repeated over and over again. That’s a fair amount of work but a lot of that is just understanding the process of how people experience your product, or your store, or your amusement park, etc.
And then there’s technology. And gosh, if you have a piece of technology that can create a 360 degree customer profile, if you can pull in all dimensions from all the modern channels, if you can use AI to sort everything into all the different types of customer engagement, if you can engage with intelligence in a personalized way, that’s awesome. So, something like Sprinklr helps a tremendous amount. Without the first two, it doesn’t work.
That’s where I think CXM is an interesting category as it evolves. I do think that some people are trying to define it as what is really customer feedback management, which is simply measuring through pretty traditional means, like surveys, and then reporting on the findings. I don’t think that makes sense. I think CXM is more like what Walt Disney did, which is you can’t just see that there’s garbage on the ground, or do a survey that people had a bad day, because the lines are too long. You’ve got to be there, and you’ve got to fix it when it’s happening. Otherwise, people don’t come back. And they tell others about that bad experience.
So, while Walt Disney got this in 1955, again, as an FYI, everyone is still figuring this out. Walt was there a long time ago. You just can’t dawdle. If someone’s having a bad time or having a rough experience or needs help, you’ve got to help them right then. So, I think any CXM platform has to not just be able to record the issues that are going on, it has to also be able to enable action to take care of the issues at the same time. You’ve got to be able to see a problem, you’ve got to be able to fix a problem. You’ve got to see a piece of garbage, you got to put the garbage away. It just goes together. And you also have to build systems around training people and building process maps, so that you can make sure this is part of your culture and the way that you operate.
So, why did I want to talk about my background. When I went to Microsoft, which was in 2006, I had the incredible honor of working for Peter Neupert, who had been at Microsoft for quite a long time, and started up the Health Solutions Group in 2006. And Peter had also been a founder at drugstore.com, CEO there. And had done many other interesting things. For example, starting MSNBC, just as an example. I also was deeply blessed to have Sean Nolan as my engineering partner. And the two of us launched HealthVault, and then worked with a lot of hospitals on a product called Amalga, which was a SQL implementation that allowed us to collect all the different data sources into one spot. So, you could have a discharge record that could then be released into HealthVault, or just sent as a discharge record.
In NewYork-Presbyterian, for example, we connected 57 different systems, and created an integrated discharge report, which was a very significant accomplishment. It took a lot of work, and would not have been possible without the amazing, amazing cooperation, and incredible partnership of Aurelia Boyer, who is the CIO at NYPH. I tell you that because that’s what I did for a bunch of years. When I first started at Microsoft, we were in Microsoft Research. We were essentially a startup inside MSR. And were funded on a somewhat annual basis for sure. But there were quarterly asks as well, as we built and grew the team and built and grew the business.
Now, I saw something in healthcare that I’d not seen before. And what was interesting about healthcare is that it’s an incredibly advanced category in a way. Healthcare really, for all the forms that you fill out by hand — and it is pretty unbelievable how much that still goes on — there’s a lot of really interesting tech that’s purchased by healthcare. All the MRI machines and CAT scanners and X-ray machines, etc. All the different robots for operating. All the different EMRs for recording patient data. Now the patient health records, the PHRs, that you can access, like Epic, or MyChart, or things that you can get through NYPH Connect. There’s a lot of systems.
In fact, some hospitals… there’s one hospital we went to a fair number of times, which is the Palo Alto Medical Center, and at PAMC they have 400 beds, so you know, a medium sized hospital. And they had 400 separate point solutions that they had purchased to monitor everything from blood pressure to the food you’re eating at your bed, etc. And managing that mess of point solutions was a living nightmare for the CIO. And quite frankly, a nightmare for everyone in the IT department. Just provisioning a new nurse or a new doctor took forever because they had to be provisioned onto multiple systems. And then when they left, they had to be deprovisioned. And that provided all sorts of dangerous holes and security issues because they would often forget to deprovision them for every system. And so somebody could then log in and get access to the health records, get access to the hospital’s systems, through some sort of rogue point solution that hadn’t been taken care of. And so, what we were selling with Amalga was a way of integrating all those systems.
But there was a far better solution that had been in development for many, many years. And it was a solution, which was a unified EHR from a company named Epic Healthcare, based in Madison, Wisconsin. Very interesting company with a super unique and very inspirational founder, who had run the company in her Birkenstocks for many, many, many years. And it created a wonderful culture out in, essentially, the middle of nowhere.
Epic basically had this idea that if you had one unified system, you would get away from all the integration nightmares that you have with separate point solutions, and you would be able to have better outcomes for patients. Because now you could see… things like sepsis, for example, where you need three or four readings with three or four different systems. In most hospitals at the time, they wouldn’t see sepsis in the early stages, because they didn’t have those readings in one spot. They couldn’t have a sepsis indicator. They would only really see sepsis as it started to present. And at that stage, often it was too late, and the mortality rate was really high.
But if you had the early detector in there, which was based on having a bunch of systems connected, then you could avoid it. Congestive heart failure, same thing. You can actually prevent a very negative outcome with salt pills. If you catch it in time. And if you wait too long, the patient often dies. You look at weight, and you look at a couple other metrics, and then you can pull this all together. We had done some really good work around that. And Aurelia used to tell me how many thousands of lives we had saved together because of these tools that we had built. But it was hard. Because we were connecting a lot of different systems, whereas Epic could do it more easily by having one system.
Now, I don’t want to underestimate the complexity of Epic. To be fair, hospitals have spent literally billions of dollars implementing Epic. A billion is not an unusual number to implement Epic. Five billion, six billion, eight billion are all numbers that I’ve heard or am aware of. But once you get it up and running, and get everything integrated, and get it unified, you’ve got something pretty special. And Epic has built a thing called MyChart, which allows you to share that data with the patient. So they built a version of HealthVault, though it was locked to the Epic system, it was an effective way. I go to Columbia Doctors today in New York, and I’m able to see healthcare data, I’m able to pay my bills. It’s actually quite a smooth system, works incredibly well.
So, why am I telling you this story? When I started in healthcare, Epic was there, but not a dominant player. At the time I exited healthcare, just a few years later, Epic was clearly the dominant player. The move had been made by hospitals to move away from best-of-breed point solutions, and to move to a unified system. And it just so happened, there was only one unified system, and that was Epic. Today, Epic is more than 60% of the EHR market. The far, far, far dominant player. It’s an amazing story that’s rarely told. It’s a fantastic investment.
So, what does that have to do with marketing? Well, when I got to the US, and I was being asked to build a martech stack, I took the learning I had from healthcare, and I said, Hey, you know, there’s a world of hurt waiting us here in martech. Because we can make the same mistake the hospitals did, and build a bunch of point solutions. Build a frankenstack. And that’s basically what everyone went about doing.
There were a few of us who built best-of-suite approaches. And when I met Sprinklr, and they talked about their unified front office, it was pretty exciting to me because they were doing the exact same thing that Epic had done. They were at an earlier stage. And there were many more hurdles to overcome, and a lot more code to write. But Sprinklr had, I believed, the right strategic idea long term. And that has proven to be correct.
So, that’s how lessons from one category or one industry can be shared with another industry. And what’s interesting to me today… we just did the interview with David yesterday. And David was talking about where he was going at Mount Sinai, and thinking about marketing as a relationship motion. And what Dave is clearly thinking about and worrying about is, how does he measure the relationship? And how does he get that single integrated profile? And I’ll make a prediction… I’m going to say that I’m betting that healthcare figures this out faster than the rest of us. And if we want to get good guideposts on how to really measure the strength of a relationship with a customer, let’s look to healthcare. Let’s see how they’re doing it. Let’s see how they’re going to think about the patient, the patient’s family, that circle of caregivers that are always around a patient. How are they identifying that? How are they managing that? How are they measuring it? They’re clearly not going to just do surveys. They’re going to have to also look at all the stuff that’s coming in… the unsolicited, unstructured data through sites like Health Grades, through all the review sites, through just general reviews that hospitals post. And all these general comments that people make across all the different modern channels, like platforms in the social area, or blogs, or forums, et cetera, et cetera, et cetera.
Very interesting times, and I’m going to keep connected to my healthcare brethren and see where they’re going with this because I do think there’s going to be some interesting guideposts ahead. And we’ll use that to inform ourselves as we go ahead together.
And with that, that is today’s CXM Experience. Hope you enjoyed that little bit of a trip down memory lane for me. I did actually have a very good time doing that. And I will be signing off now. So, I’m Grad Conn, CXO at Sprinklr. And I’ll see you… next time.