PODCASTS/HEALTH PLANS/TECH: Interview with Stan Nowak, CEO of Silverlink

Here’s the transcript of my interview with Stan Nowak of Silverlink. The original audio is here.

Matthew Holt:  Hi, it's Mathew Holt of THCB and I'm back with another podcast. And this time I am talking with Stan Nowak, who is the CEO and President of Silverlink. You may not be familiar with Silverlink; they are a company that relatively new and relatively small. But they have a big client base doing, primarily outbound automated phone calls. We spend a lot of time on THCB talking about the Internet, the web, Health 2.0, and all these kind of new ways that people are using computer technology to communicate with each other. But what we obviously we don't realize, well we do realize but don't think much about it, is that in healthcare; a lot of the activity is in phone calls. And there is a big market in trying to do the phone call better and to gain more information from that.

So, I chatted with Stan a couple of week back and I thought it would be interesting have him on a podcast because Silverlink has introduced some new initiatives. But they also have been doing a lot of stuff that probably is not familiar with people. It has sort of been under the radar, both in terms of working for PBMs if you have Medicare Part D and working with many health plans. So, first off, Stan, good afternoon. How are you doing?

Stan Nowak:  Thank you very much Matthew. I am doing quite well today.

Matthew: Quite, quite well. Stan lives in Boston, so he is not doing as well as those of us on the west coast. That is just a THCB impression. So Stan, given that I am a mere dumb blogger, but I'd never heard of Silverlink, I suspect that you fly somewhere under the radar. And perhaps interviews like this are hoping to sort of poke your head up a little bit about it. But you very much are involved with kind of something that seems to be sort of in the guts or the plumbing of healthcare organizations, people don't think much about, which is helping automated phone calls. So just, tell us a little bit about that business and what Silverlink does and brings to your customers in that business.

Stan: Absolutely. Yeah, Silverlink is in the business of providing automated voice solutions, exclusively for the healthcare industry. And as we mentioned the other day, we enable our customers to proactive deliver and hit the client and personalized interactive phone calls that are automated, but are recorded in a human voice and navigated by the call recipients voice. And they tend to be, across the experience that we have had over the past five years, much more effective than letters, email. Often much more effective than the web. And clearly a lot more affordable than using live call center agents. So for a certain category of communication activities that health plans and pharmacy benefit managers and DME suppliers undertake, this is a very effective way to reach out proactively with a personalized message and get very high response rates around a range of problem sets. And the problem sets are similar to the ones you are seeing and talking a lot about with respect to the web, you know, content development that has been going on for the past several years.

And as you well know, with the migration of healthcare toward a consumer model, as many folks have said by now, you can squint a little bit and see healthcare going from a B‑B business model to a B‑C business model over the next few years. And that creates a great deal of urgency for members and patients to make decisions that impact them financially and that have, obviously, many other implications from a quality standpoint. So our business is really helping members navigate a complex healthcare system.

And for the past five years for, as you point out, seven of the ten largest health plans in the country, and many, many other companies in the healthcare area, we've been helping them do about a 100 different types of things that all have the common theme of either trying to drive behavior, whether it's medication compliance or enrolling in a particular type Medicare Type D plan. Or getting a mammogram screening, or whether it is collecting data from members, or educating them. There are a lot of education programs around. You know, “what is the Medicare Part D benefit for seniors.”  So, we have been doing that for about five years, and again, the problem sets are the same problems that are being addressed by many media. It just turns out that with the demographic of healthcare consumption, just as one point, seniors are 13% of the population and consume about 42% of the drugs. You know, this is a population that is becoming more adept at new media and web and email. You know, it is certainly very effective to contact them by the phone.

Matthew: No argument from me on that one. If you look at user rates of the web, for those over 65, they are dramatically lower than in the working age population, and that has always been the case. So, let's get down to the sort of nitty gritty. I am a, let's say I am a Medicare Part D potential recipient, or I am a retiree with a PBM administered benefit or whatever. I am sitting at home and the phone goes, I pick it up, what happens next?

Stan: Immediately, you don't hear any of the day that people typically associate and find offensive with calls. Immediately when you pick up the phone, it says, a recorded human voice says, 'Hello, this is USA Care with a call for Matthew Holt, is this Matthew?' And you say yes or no. So immediately, first of all, what you see on the Caller ID, this is your healthcare, in many cases your managed care company calling you. That is what appears on the Caller ID. When you pick up the phone it starts speaking to you. It mentions your name and then we authenticate who we are speaking to. So we may say, 'Is this Matthew Holt? Yes or no. For your protection and privacy would you please say the last four digits of your Social Security Number? One, two, three, four.' And then we deliver the body of the message once we have reached you, you've answered, acknowledged that you are who you are authenticated, then the body of the message. Whether it's 'according to our records your prescription for Lipitor is about to expire, would you like to refill it now?' Or whether it's, 'we understand, based on our records, we understand that you are on the drug Vioxx, you may have heard the drug has been recalled earlier today.' You know, those kinds of things.

So, and then you answer, you could answer a series of questions. We could be doing a health risk assessment where we ask you questions about family histories and things like that. And your answers are recorded and based on your answers we might as you a different question or transfer your call to a live call center agent, or if it was a clinical program you might say something about your blood pressure that would lead us to send an email to a nurse care manager at a disease management company.

Matthew: Right. OK, so you are heading down a series of paths here. So just on the basic mechanics, did the responses come back, is it mostly voice recognition or is it mostly people hitting a one, two, or three? You know, press one for this, and press two for that.

Stan:  It's all voice recognition.

Matthew: So, it is all. Speaking of somebody whose accent doesn't get fully understood by every single voice automation system that I speak to in this country, I don't know why. [laughing] Let's ask two sort of obvious questions, first is for the voice recognition: how many people... How does it deal with different accents, different languages, and all that kind of stuff? And then there is a macro question which is what response rates do you get to people who A, actually acknowledge and pick up the phone and put in their information or identify themselves and then B, make it all the way through the calls. Those are two different questions.

Stan: Yes, absolutely. And I think the issues with voice recognition, first of all, I said it is exclusively voice recognition but there are defaults. So if the system doesn't understand you, it will ask you to repeat and then we will in fact default to DTMF or pushbutton. You know, press one if you meant to say five and two if you, you know. So, you default to numeric inputs. So, the first question... The way you get voice recognition accuracy, and I'm not going to tell you that if you speak a paragraph voice recognition will pick that up and accurately transcribe that into data. That's not; we are not there yet with voice recognition. What we are doing is asking a series of question and limiting the number of answers we are listening for. And that dramatically improves the accuracy of voice recognition as a technology. So if there are five potential answers we might list those answers and you are going to be saying one of them or other.

Matthew: I'm with you, so you are bringing that in from Nuance or some other vendor I assume is actually a part of the technology that drives that part I assume.

Stan: Absolutely, we embed best‑of‑breed voice recognition technologies and there are also text to speech technologies that take data from your doctor's name and read it in the call for example. So that's sort of the first answer, is that... given the way that you structure these calls, the voice recognition aspect can be extremely accurate with very, very low error rates.

The second question is really the more important one which is response rates. And that's really what... when people are balancing the types of modes of communications they have to drive behavior, educate or capture data from populations, they have to think very hard about what's the response rate. They also have to think about the core attributes, is it HIPAA‑compliant, is it a pro‑active mode of communication? Is it digital? Is it cost‑effective?

But with respect to your question about response rates, that was something that surprises, pleasantly surprises, all of our clients when they start using the programs. So for example, if you're doing... you care about a few things, you care about response rates, who listens and who does what you're asking them to do on the phone call, but the more important question is who takes action based on the information you're delivering? So for example, in a Medicare population for a simple call which was a flu shot reminder, and flu shots as you know are important for many members of the population, but what people want to know is 1, the time to get it, 2, should I get it and 3, where can I get one based on my address? So what we do are calls that explain and use... and I know you know Don Kemper and the Healthwise organization... we have a partnership with Healthwise where we are delivering clinical content, peer‑reviewed content that is Healthwise certified about the benefits of the flu shot, why you should get it, answering questions about it, and when we reach out, to say, a Medicare population and we'll have a reach rate of 73%, or in the 70s, a very high reach rate, based on both the outbound phone call and then people who call us back to get the message, than they would have received if we had have reached them on the phone.

And then the question is how many go and subsequently get the flu shot itself, so in a... and we get various data back from our clients, in populations we've seen as much as 40% of the people who are reached with that kind of call go out and get the flu shot, which is a very high response rate. I'd say our response rates... you know, we do a hundred different things, response rates vary based on the type of message we're delivering and who we're delivering it for... and we get response rates from 5% to 80%, so there's a fairly wide range, it's not a simple answer but the response rates are typically much higher than alternative modes and much faster. You get the response within two or three days rather than waiting six weeks for 15% of your surveys to come back.

Matthew: I'm with you. So... do you happen to have, maybe an unfair question, but you say there was a 40% of people who actually went out and got the flu shot, do you have an idea what the control group was for the other people in the health plan that didn't get the call? Have you done any work like that?

Stan: We do that... we have various levels of success in tracking those measurements with our populations and we do that, and we have some data typically they're measuring what is the cost of the alternative mode of communicating and so for our flu shot reminder we'll have a client tell us that it was as effective as using humans and saved them 65% of the cost. For health risk assessments it was two times as effective as their alternative mode of delivering those health risk assessments and it was 50% less cost. So results data is somewhat spotty, we do get it and our customers are consistently achieving higher results faster in digital form and doing it at a fraction of the cost and in a fraction of the time.

Matthew: Sounds good. So that kind of leads me to the next point which I think is the business model. You're going after people who are harder to get other ways, it's cheaper, faster, easier... you're basically introducing the option to a customer of both doing something they either weren't doing very well or were doing very expensively and doing it cheaper, and getting some good results out of it. So that's nice. But I guess the unexplored opportunity here is the disease management/outreach/outbound customer service of any kind for health plans has basically been done a) not very well, that's a quiet thing we shouldn't be mentioning too loudly, but b) being concentrated in a very small number of people. I'm a typical health plan member jumping from health plan to health plan, have typically crappy experience with them when I actually need something from them, have to deal with claims or whatever: there are a lot of errors that go on in that process and that's very typical I'm afraid for most health plans.

But more importantly is that other than getting an email maybe once a month from my health plan, and by the way I still get emails from the health plan a I left a year ago interestingly [laughs], I've been getting nothing...well my new one did communicate with me recently to tell me my credit card was about to expire, so I guess they care about the right things [laughs] But other than that I don't hear a whole lot from them and you can argue back and forth whether people want to hear from their health plan or whether they want to hear from their provider groups but the answer is there's probably things that even an overweight, middle‑aged male like me should be hearing about, and should be engaged with, about health that we're not, that if you can get there cheaply, may actually have some impact down the road.

So I guess the question is, I assume when you say you're going to the disease management, you've done this thing with Healthwise about diabetes, and that that's the group below the waterline in the iceberg group that you think are the ones who you can get at. So tell me a bit about, if I'm going in the right direction here, and also give me some of the specifics about what you're intending to do with the new diabetes program.

Stan:  Right, well there are a couple of points here and I know you're familiar with the predictive modeling world.

Matthew:  Familiar is a strong word, but I've heard about it... [laughs]

Stan: But increasingly we're hearing more about population health management rather than disease specific management and the perennial challenge is not only how do you deal with your high cost populations which you are typically best going to do with nurses and care managers on the phone with patients. But increasingly people are concerned about the at‑risk populations, the people who are not high‑risk now, not high‑risk and high‑cost, but at risk to becoming high risk and high cost and in the diabetes world you've got 21, 22 million people with diabetes and you've got about 54 million people who are in this at risk category.

So the real question is with those kinds of populations suffering from chronic disease or at risk for suffering from chronic disease, when you think about the population of nurse care managers available in industry to care for those folks on the phone, the numbers don't... there is a nursing shortage and there is not enough capacity in the nursing and care management world to effectively care for them, never mind considering costs.

So there are two issues here, one is how do you most effectively deal with the high risk, and the challenge there is effectively utilizing your high value clinical resources: nurse care managers often. And then the other question is, how do you keep people from becoming high risk that are at risk, and to date there's not a lot that's done there, maybe letters are sent. And predictive modeling allows you to know who these people are and identify the populations, but the perennial problem is, there's no cost effective delivery mechanism to get people to take action or to care for themselves or to avoid the events that occur that turn them into high risk populations.

So what we've done in disease management and are continuing to do is build a life cycle of engagement with the disease management... with a population, and map the types of communications we deliver both to the workflow of the nurse care manager as to trying to engage the coach and monitor the health of the patients that they manage, and also the life cycle that a patient goes through the course of a year, provide them with educational resources and measurements so that we can monitor and track their behavior and their progress over time.

So, essentially, that's really it. We're working with the nurse care manager for the high‑risk populations. For the at‑risk populations, there are programs that are delivered to those populations that are educational in nature, to keep them from becoming high‑risk.

Matthew: But you're also gathering some data on the phone from those populations while you're doing that. Obviously, you can measure that, yes, somebody sat through a lecture on the phone, or an explanation on the phone about eating better, taking exercise, drinking less, or whatever the 'behave better' lecture is. But also, you're taking measurements from medium to high‑risk people on the phone, as well? Are you asking them to put in measurements around blood sugar, hemoglobin levels, that sort of stuff?

Stan: Yes. Absolutely. Those kinds of readings and measurements, reporting them on the phone, and comparing them to what they said last time and identifying increases or when measurements are moving in the wrong direction. Based on those clinical increases, we may be taking action on the phone, transferring them to a nurse at that point in the call if we see measurements going in the wrong direction. We may be emailing a specific care manager who is assigned to that patient, to provide them with information that they will then use to follow up with that patient. That's exactly the type of thing we're doing to monitor and help manage that population.

Matthew: Which leads us to another issue, which I think is much more complex. I think, as this type of activity grows...and we'll discuss in a minute whether it's in the best interest of health plans to be doing this type of thing in the first place, to become disease managers. That's a bigger argument. But for the ones that are going down this path, I have a separate article I call "The Yin and the Yang of Health Plans", which is that they have all these smart, medical directors inside who want to do all this better healthcare management. Then they also have a bunch of actuaries who want to get rid of all the sick people and have no interest in being better healthcare managers.[both laugh] But, given that you're on the side of the angels and working with the folks who are trying to do the right thing in terms of the better disease management. They've got this other problem, which is that there are other sources of that data.

I'm thinking, just on the health plan end, that some health plans now are picking up data from lab tests other avenues. Some of them, not a lot, but some of them are now starting to work with either provider groups or directly using interventions directly in the home—things like collecting data from visiting nurses or they might have an automated tool like Health Hero or equivalent.

Stan:  Health Hero, yeah. Sure.

Matthew: So there's other data coming in the same kind of vein. Then, of course, there's people going to the web (there aren't many, but there are some) and inputting their own information into their own peer chart or a place the health plan may have access to. As we go down the path, this data from different sources is going to become a bigger issue. Is this something you've run into yet? If so, how do you think health plans are dealing with this? Are they going to throw up their hands and give it up and say, "We're going to have to have a separate organization that is our data cruncher." Is that were you guys are going to end up? You not only have the outreach part, but you're also collecting the data. So, give me a sense about how you think about data integration of all this stuff that's out there is going to happen.

Stan: Well, I think that clearly, the more and more accurate data you have access to, the smarter you can make the interaction. Just think about the sort of information you'd want your nurse to have access to when they're on the phone to a high‑risk patient. You'd want access to any and all information that's relevant to that patient's care, and as up‑to‑date as it can possibly be. The same holds true for an automated phone call. Our interactive phone calls can be highly sophisticated, and can bring in data. It can bring in numeric data. It can bring in your doctor's name, the clinic you went to, and the date of your next appointment. All of these things, this data that can be brought in and discussed in the phone call. Clearly, the more information you have access to, the more intelligent the call can be. We are firmly in the world that everybody else lives in, which is that data is an imperfect place right now. It will be for many, many years. In the meantime, we're doing as well as we can with our clients in terms of their access to data. We can't be better than their ability to provide, or provide access to, the most real‑time, up‑to‑date data on these patients. But we can be as good. We can take whatever they have access to, and build that into more intelligent, sophisticated dialogues with patients that are more relevant to them.

Matthew: So, you're kind of at the mercy of what your clients have. Is the majority of the data that you're integrating into current calls; is that coming from a claims system, or some kind of scrubbed CRM system? And, give me a state of the industry, as to how good that data is. If it's getting to you in the right place, it's a decent assumption that it could go to other places, like personal health records and other kinds of consumer front‑end applications. I'd be interested in hearing where you think the world is, in terms of data integration between all those different silos and health plans. I don't want to insult your client. [laughs] But I like the idea that they know where they are.

Stan: My sense of the situation is that, obviously, it varies by plan. I think generally, we are young in that area, of having very timely information that's available in a centralized place, in order to take action based on it. I think we're young, in that area. That's going to be an evolutionary space for the next several years. To get much more detailed about that, I'm certainly not the one who is highly qualified to comment on that in my organization.

Matthew: You're suggesting that CEO is not the guy who stays up with the Access database until four in the morning trying to figure it out [both laugh] Yeah, I can believe that. It's good to be the king, sometimes.

Stan:  [laughs]

Matthew: Well, moving on from that, let me just nudge you one more about a part of that. Do you think that is a core element of what Silverlink does? Because I'm dealing with some companies that are dealing with health plans and are trying to figure out front‑end integration of websites and back‑end integration of data.
I've done some work, over time, in the back‑end integration of data from different sources like the health plans, the drug plans, and elsewhere. It's always, as you said, dealing with data is always a mess. You get the impression that some businesses, not necessarily in healthcare, but outside our business, there are some industries which have now really figured this stuff out. They are now able to move to a data‑centric view of the customer, which can be moved into the other applications it needs. They have all the flexibility and can do that at a lower cost than they've done before. They're making money out of the ability to do that because they start figuring out different ways to serve their customers and generate more revenue out of it. We've been talking about this in healthcare for awhile. Do you see that as a business that Silverlink would ever be in? You kind of are around the edges of it now, but is that something that you may actually get in, given that you've got all these capabilities of introducing data within your system as it is? Or do you think that's something you're going to stay away from?

Stan: I absolutely think that data management is something that is part of our business. Clearly, if you step back and think of the larger context here, the real goal is to deliver the right information to the right person at the right time. I was saying that for some time before I realized that Don Kemper had said that 25 years ago.

Matthew: [laughs] I think Keynes said something about how every genius is actually parroting the writings of some defunct economist, and Don isn't quite defunct.

Stan: [laughs] So, I thought it was quite clever until I realized I'd just taken it from his book. Really, when you think about the challenge that specifically, health plans have, health plans have a credibility gap with consumers. No one is claiming that health plans have redefined consumer delight. However, it is a major industry challenge for health plans to assert a position of trust and redefine the way healthcare consumers think about those organizations. In my view, and not only my view, one of the ways they will reestablish a reputation with consumers is by taking information that's both required of, important to, urgent to an individual at a time when they're making decisions and offer credible information and guidance to that individual at that time.

So, to me that is looking at an event in the claim system that says, "OK, this person has had this kind of a diagnosis, and that means they're probably asking these types of questions about treatment." Now, I think we're a long way from the health plan telling you what treatment to pursue, but we're not a long way from the health plan saying, "We understand that you have recently been diagnosed with this. You may be trying to find information about the following subjects. Let us tell you what the information resources are that are available to you, among others, as you explore the options that you need to be thinking about right now."

So, obviously that requires a high degree of data integration and it's real‑time. But, clearly health plans have access to the source of information that can allow them to time their delivery of information and decision support to an individual at a very important time. I think it's critical that they take advantage of that asset to both help people do this navigation, it's becoming more urgent to them and there are more decisions and risk associated. But use what information assets they have to time those communications. And again, redefine a relationship of trust with the consumer. I do think integration is part of our job. It's part of healthcare's job. And it will help healthcare, specifically the managed component of healthcare, redefine their relationship, which is a priority for managed care. And you are absolutely starting to see companies recognize what they have to do to get there.

Matthew: That leads us on to a segway, a little bit, into where you think your business is going, because you obviously rationally take a look at the world and think, "Who's got the biggest problem with outreach, and who has the most data and can do something about it?" Health plans are a great place to start. You told me when we spoke the other day, the Silverlink, it started with a different idea to where it was going to be focusing? I think you mentioned it was, I'm blanking on that. Was that going to be in the home?

Stan:  Yeah.

Matthew: Anyway, the upshot is that one of those things, speaking as someone who is not a great fan of the current setup of how HSAs and CDHPs have been marketed, but is a great fan of improving the experience the consumer has with the healthcare system. It doesn't matter to me, or I think to most consumers, whether you're getting your information and your services from a private health plan, Medicare, the government, or the NHS in the UK, or a provider group direct, or a big provider organization, whatever. You want to be treated like a human being and you want to have good information in a timely way, and you want good customer service from whomever you're dealing with.

Any organization of any kind that's delivering healthcare services of any kind has got to wrestle with that. As we said, I think across the board, whether it's private, public, here or everywhere else, they've all done a piss‑poor job. Things like Silverlink are helping; they've sort of bridged that gap a bit. So, having said that, you're obviously focused on the health plan market at the moment. Do you see opportunities elsewhere? How far are you down the path, talking about providers, talking about international, talking about taking this technology or service to other areas?

Stan: We certainly work with the managed care complex, and within that I consider the PBM area, the specialty pharmacy area. We also work in the medical supply world and a few other sectors of healthcare. Again, exclusively healthcare, and the common theme being those people who have to manage populations in the home and are interested in their behavior and education, and all that kind of thing. We are exclusively within healthcare because of the number of issues: clearly with this transformation from B to B and B to C that's going on right now, and whether it's a cause or effect, the commercial models and the plan structure that are creating incentives and urgency for individuals to better understand their healthcare and better manage their healthcare consumption. The opportunity area in this world is perhaps not limitless, but certainly there's a tremendous amount of headroom in the marketplace.

To that extent we are interacting with providers. And integrated delivery systems, certainly. Our issues are really just making sure that, again, as you point out, a common theme is anyone who is interested in driving behavior, capturing data, or education populations, regardless of whether they're a managed‑care organization or a large clinic. People who have large populations are potential customers. Silverlink, the core innovation is that we've removed the barrier to utilizing this technology. We've allowed people to build call programs to call a thousand people to drive specific types of behavior. That's economic, and it's able to be set up in hours, as opposed to traditionally set up over several weeks and you need special software developers to do it.

So, we built the company around a concept which drove us to have a very intuitive way to build interactive dialogs that can be extremely sophisticated but can be built and delivers to a small population. There really isn't a limit to the types of people that we can work with within the healthcare industry. We've five years‑old, so we're still very early as a company; we're an emerging‑growth company at this point. We see the opportunity as almost limitless in this market.

Matthew: I'm sort of listening between the lines of what you're saying. Does that mean that you think that the health‑care PBM managed care has got a lot more opportunity with Silverlink without you having to go much outside that for the time being? You're not saying you're shutting yourself off, but I think that was what I was hearing. So, am I right?

Stan: That is absolutely the case. I think what's important for folks to understand is, people talk about consumerism and consumer healthcare, and people make the mistake of thinking that CDHP or consumer‑directed health plans are consumer healthcare. In fact, that's a plan type, but that's a very group product. Employers choose a CDHP plan and then they put their employees into it. The more intensive consumer experience, and MacKenzie did a paper recently on this, when you look at the Medicare plans, the Medicare Part D plans that have come out, those are truly consumer products. Individuals signing up as individuals. They can walk out the door at the end of the year, and walk to the plan next door. I think that's really been a wake‑up call to the industry with tens of millions of people in this new consumer health‑care product. It's exposed the cracks in the consumer delivery services for health plans. It's created an entirely new range of education around what it means to interact with a high‑consuming population with a relatively complex plan.

That's a couple of years old. We're very early in having the lessons of what consumer, business, and healthcare means to health‑care companies. Because of that, we're really at the beginning of what I think is going to be a dramatic change in the way the health plan complex interacts with consumers and we're very early in that. So, I think the opportunity right now is really massive in that area as we undergo this transformation.

Matthew: And I think that is dead right because the more you put the onus of certainly plan selection, but even things like the selection between benefits, this, that and the other onto consumers, let alone people start thinking about ones who are sick or may be sick. You have a huge amount more discussions and choices. And people at the moment again, that is done very, very poorly. And you are in a great segment there.

The last question where I let you talk a little bit about the company. Are you hearing from your customers the desire that you should also be representing them with, kind of, other modes of getting at their customer? Obviously you are in some extent competing with mail, and email, and web based platforms. Are they saying to you, 'Wouldn't it be great if we can have a one stop shop and you could provide the web stuff as well” or do you not hear that... Are you pretty much satisfied to be best in class at what you are doing at the moment?

Stan: Well, I think what we do at the moment, the automated interactive calls and the information systems around that, is something that clients certainly don't have a competency in, and don't have staff currently doing. Every one of our clients has a significant web infrastructure; they have made large investments there. They have tried email initiatives. And some of those are obviously successful and ongoing for certain populations where they have the email addresses, and where they are not delivering protected health information in emails. So what we have really started by doing... Is delivering something that they just didn't have the competency or didn't have the infrastructure in house, to do in any scalable, high quality way. And, but your point is well taken. As we develop more and more information around what is the best way to drive behaviors to populations. And as populations require information more than having information pushed to them, you know, as there's more urgency around information, you think about your mode of communication a little differently. Because there will be a pull for the information rather than a push.

Right now we are still a little bit in that push mode because the interests are more heavily on the side of the provider of information than on the consumer of information. I think that is changing, I think it will change in the next few years. And that allows you to think a little more broadly about the modes. Something we are absolutely thinking about and a lot of the core intelligence that Silverlink builds about how behaviors are best driven, is mode agnostic. So, yeah, I think you will see as we move into the next few years, that Silverlink will provide the current disciplines but will be more mode agnostic over time.

Matthew: Well, that is interesting. So, just to wrap up Stan, give me a couple of data points as to where you guys are in your development as a company. You mentioned last time you had around 80 people and awhile back you raised around $14 million in venture capital. Give me a sense of your growth rate, any numbers you can tell me about, kind of, you mentioned some numbers around clients. I am just trying to get a sense of where you guys are in that kind of growth cycle and what you are looking towards.

Stan:  Oh, absolutely. We have actually have a little over 60 people now but we are growing at 80%.

Matthew: Sorry, I got my numbers mixed up then. I'm reading what I wrote down last time and as most people know, I was writing down what you said, that doesn't mean it's correct. [laughing]

Stan: But we, you know, in '06 we grew about 80% over '05. We have been around for a little over five years now. And we have about 45 and 50 customers. And those are really in the five segments that we talked about: disease management, manage care, pharmacy benefit management, and durable medical equipment or medical supply. And, so the pace of growth of the company is obviously a, we are a high growth company. We have, we are doing over 100 different things for health plans and all the segments of our marketplace. And that grows continually, so it really is an unlimited, you know, the problems statement of 'when you need to drive behavior, capture data, or educate populations' you can imagine how broad that problem statement is. So, obviously we are continuing on a very high growth path.

And I think the wind at our back is the trend toward consumerism and the transformation of health plans, or the healthcare system from B‑B to B‑C. And those are sort of the macro trends, and in the meantime I think we are also seeing with respect to automated voice communications, people have been somewhat resistant to that over the past five years. And we had to do a lot of missionary work in the early days. Today, there is much, much less resistance. I think based on the fact that there's been a tremendous amount of operating success across many, many programs across large populations. So we have called, we call tens of millions of people a year on behalf of our clients. And it's continually driving high impact, and very high quality results. And people are not complaining about these calls. They think about healthcare calls very differently than they think about telemarketing calls. And that is really critical. And frankly people do pick up the phone when their health plan calls. They generally see that information as having much greater importance to them than other types of calls they might receive at home.

Matthew: I think you are right. And especially given the amount we get called these days on Election Day. I am always called on Election Day by about 27 famous politicians and the only issue is I'm a mail voter so I've already voted two weeks before. [laughter]

Stan:  Well, you know what? I'm in Boston. We are expected to vote more than once of course. [laughing]

Matthew:  Well, I guess with coming to Chicago you couldn't believe it even more. [laughing]

Matthew:  And finally, when did you raise that last VC round? When was that?

Stan: We raised money last spring. We have done three rounds of venture investment. And we are not looking for; we are not looking for more investment. We are in a... We are high growth but we are also not looking for more private investment at this point.

Matthew: Well, I am just being cynical here because you can't say this as the CEO, but I can. But usually when there's VC guys giving you some money at some point they want it back for some reason. I don't know, actually they want more back. [laughing] And that typically means at some point that there is what's called a liquidity event one way or the other. I won't ask you how close you are to that, but...

Stan:  It is unusual, our guys said, 'Here is the money. You know, keep it.'

Matthew:  Yeah, keep it.

Stan:  I thought it was unusual too.

Matthew: Well, there was that guy in 1990 who took like $27 million from VC for a fake company that had a fake video screen or something, I forget what it was. He had “The Who” play at a big party in Las Vegas for him and all the rest of it. And awhile later they noticed that there was no technology, there was no company, and woops. But that kind of stuff isn't happening quite as much anymore sadly—probably happily, actually. [laughing]

Stan: You know, I've got, I think what's... I am fortunate, I have very high quality venture investors who see and share our view that we are at the beginning of something that is very, very big. And that has tremendous potential over the next several years. And, so we're here and we're building a great company.

Matthew: That's great. And then finally of course, the last question I have the ask you is you raised this before we went on, before the recorder was on, is that, you now have, I'll give you 15 seconds to defend Reggie Herzlinger, Michael Porter, and Clayton Christensen and any other Harvard Business School graduates who I may have defamed on THCB at any stage.[laughing] You don't have to answer this.

Stan: I think that all I will say is you are welcome on behalf of the institution that provided the solutions to the healthcare issue. [laughing]

Matthew:  You mean on behalf of Stanford? [laughing]

Matthew: All right, OK. I have been talking with Stan Nowak, who is the President and CEO of Silverlink Company doing automated voice outreach, mostly on behalf of health plans. And we had an interesting conversation about the current status of his company and the future likelihood for more automated voice calls coming from healthcare organizations. And I suspect the answer is there will be a lot more and you can all be expecting to get one when you are sitting at home watching TV or having dinner anytime soon. So Stan, thank you very much for your time. It was nice talking with you.

Stan:  Thank you so much Mathew.

Matthew:  All right, take care now. Good bye.