eHealth, Inc. (EHTH) Earnings Call Transcript & Summary

All right. We will gets started. Hello, everyone. I'm Jailendra Singh, healthcare technology and distribution analyst at Crédit Suisse. Thanks, everyone, for joining us. Next up, we have eHealth management team for a fireside chat conversation. From the company, we have the company's CEO, Scott Flanders; CFO, Derek Yung; and VP of IR, Kate Sidorovich. Thanks a lot, everyone, for being with us. I'll begin with some of my prepared questions, but if anyone has any question, please e-mail them to me at [email protected], and I'll ask on their behalf.

Scott, Derek, I mean, we have talked about this several times, but for somebody who's listening to the story for the first time, maybe take a step back because there have been some confusion recently, and people are thinking that the whole market is broken, the industry TAM is not real. Let's take a step back to begin, share with me your thoughts around how you see the top line growth over the next few years? Maybe spend some time on the key growth drivers for your business? And has anything changed with those growth drivers because of the company's experience around churn rate recently?

Right. So the overall market is enormous, $62 million eligible in Medicare. The population continues to age, people are living longer. So more coming in the funnel than are going out of the funnel every single day. We then have, on top of that, an important mix shift. So everyone who's over 65 is eligible for Medicare, and they don't have to do anything, they're just eligible. But if they wish to, they can buy a Medicare Advantage plan and that's the business that we're in, selling Medicare Advantage plans, Medicare Supplement plans, and those offer benefits well beyond traditional Medicare. And it's a true win, win, win. It's win for the beneficiary because they have better patient outcomes. They're healthier. They have a primary care physician. They have a general contractor who looks after their health, make sure they're coming in for their regular checkups, making sure that they're drug compliant, making sure that they have the right specialists for any condition that emerges. And so it's better for that person. Many of them can join in these plans at 0 premiums, paying no insurance premium. So you say, well, how can that be? It's almost too good to be true because the carriers make a lot of money on these plans. And so they're very happy with it, and they've been wanting to grow their business, and I'll come back to that. Within the government who reimburses the carriers on a per member basis, the reason they are -- have been heavily supportive of Medicare Advantage is because it's cheaper for them on a per beneficiary basis by significant amount of money than on the traditional fee-for-service, which is an all-you-can-eat buffet style of care. And so it's not as effective and it's more expensive. So this is really a case of it's a win of the beneficiary. It's a win for the carrier, and it's a win for the government. Our little space here, which we think is going to be billions and billions of dollars of revenue and many multiples of that of enterprise value, we play that interconnection between the beneficiary, the senior and the insurance carrier. Now they have their own distribution systems. Those distribution systems are not their core business. They're not inherently efficient at it. It's an adjunct to their core business of delivering care and assessing risk. And so for us, we have been helping them grow their business. Right now, Medicare Advantage is projected this year to grow 10%. If you look at the total telephonic channel of which select go health at par, we're growing at far multiples of that growth rate. Last year, we grew Medicare Advantage 100%. And so this channel has become increasingly important to the carriers. Then you add this year with COVID and the resistance that seniors have to take in-person meetings and appointments is going to just accelerate that migration. That was already multiples of the underlying growth rate, and we believe it will reach an inflection point. And that's being experienced by all of us. That's why all 3 of us staffed up our agent counts quite dramatically. It's why the -- all the carriers are supporting this channel with soft dollar, advertising dollars, sponsorship dollars to try to accelerate the growth because they realize they're becoming increasing reliant on our channel. So it is just a great sector. I can't think of a sector that it's a better place to be because we're helping seniors get into the right plans, and those plans make them healthier at very minimal cost to that. So what could go wrong? Well, what went wrong for eHealth last year? Though we grew 100%, Jailendra, some of that growth was unprofitable. And we didn't know that at the time. The assumption that we had was that there was an average lifetime value from all channels. And as we were growing so fast, we put all of our energy into maximizing that growth and improving the conversion rates of the leads that we brought in and the fulfillment source for those leads. And that optimizing around that metric drove some unprofitable enrollment. So how did that happen? I think it's important to understand the root cause of the churn, which is largely isolated eHealth. It's not something that's toxic in this entire sector. We grew faster certainly than at least one of our competitors. We grew very fast. And a couple of deciles of that growth, as I said, were unprofitable. What was happening? By optimizing around conversion, we were spending money in channels, primarily direct response TV and at times in media spots that were driving calls coming in with seniors would have high propensity to say, yes, to whatever we were doing for them. And then we were routing those leads to agents in our workforce that had the highest propensity to close a sale. And so by doing that, we were bringing in heavy switchers on the margin. And so that's what elevated our churn. We did not realize this. Some churn we knew was going to come because of the OEP and then the extended SEP that was COVID-related into Q2, so up maybe 100 basis points, 200 basis points. But we never anticipated, did not see it coming and perhaps were subject to legitimate criticism on that, but we didn't. As soon as we saw it, we unpacked it, we went to the root cause of every 100 basis points of churn. We realize that 85%, 90% of that -- of the factors were not marketplace driven, but were internally driven. And so I'll let you ask specific questions, but it was really in the early summer that this emerged as what we saw as a critical issue. We attacked it, we took initiatives. And we're very confident that those initiatives are on track to restore our LTVs in 2021 back to 2019 levels.

Let's talk about those initiatives. Maybe provide an update on all those initiatives you've been putting in place to improve the churn rate? And how -- I mean, are you going to wait? And when are we going to see the results in your numbers? I mean, of course, you can provide a qualitative commentary right now. But when exactly, I can see in numbers, hey, yes, these are really making impact on your churn trends?

Right. And so the churn had elevated in Q1. But because of historic seasonal moderation, we did not see it as an alarming number. When it -- when we reported Q2 and had 42% trailing 12 months, we had already recognized it and attacked it. And in Q3, the churn seasonally stabilized. There were no -- there was no impact of any of our initiatives into Q3. We now fully have our major initiatives. So the first one is just fulfillment source. And so we saw that we had not only lower conversion, which we've always known, and we've actually improved the conversion from our external agents, but we had significantly lower LTVs from our external agents. So this year, we've increased our internal agent hiring by 120%. Our mix shift has gone from 30% internal agent capacity to over 45%. And so we'll bring in far more -- even last year at 30%, we had 50% of our enrollments were internal agents. So we do ramp more leads to our internal agents. The other piece of this that's very important is we have much lower churn, 25%, 30% longer persistency with online enrollees. And so we've significantly increased our fulfillment mix from -- to online. In our Q3 print, we guided up to 45% to 50% of our enrollment anticipated to be online, assisted and unassisted combined. So those 2 factors are significant. The second factor, big bucket, would be our changes in sales processes. So that was -- we're now optimizing -- we're sorting leads, and we are allocating leads based upon profitability instead of optimizing on a singular financial metric of conversion. So we're taking into consideration projected LTV. And we have that information on a look-back basis. So a much more sophisticated lead segmentation and lead allocation. Importantly, agents are coin operated. Look, that's -- you hire them because of their ability to gain trust and confidence among seniors and close a senior on a single call. And they get paid a commission for doing so. We were paying our agents 100% commission on approved applications, okay? We realize that, that was a flawed approach. And we were slow to align agent incentives with what is in eHealth's interest. So today, for all agents, internal agents, we withhold 30% -- on average, 35% of the upfront commission that we were ordinarily paying until the 100-day point. So that -- most of the churn, a very high percentage of first year churn, as you know, happens in the first 90 days before they become a paid member. So that will be changing agent behavior. And then the third large bucket is just our customer engagement and retention efforts. And so we, again, set up a smallish retention team started out around 30 full-time customer support agents. And the call volume was very high. The retention was very strong. Upwards of 90% of the calls we answered, we were able to either answer their questions or retain them in some other way. And so we have significantly increased headed into the busy AEP period, our retention team, upwards of 200 people with enrollment specialists and customer care. So these 3 factors together, we're confident will have a meaningful increase. I mean, the number that we've set, we're going to -- we have a 6% hit to our LTVs in '20. And we've said for '21, we'll minimally get back to '19 levels. That only requires a 7% improvement. And as you know, we're getting a 5.5% commission increase next year. So all of these initiatives, on the full year, only have to improve 150 basis points our retention to get us to that level. Well, that seems very, very conservative, and our internal objectives are much higher, but that's what we committed to. That's what our guidance and budgets will be built off of.

Okay. Just quickly a couple of follow-up on your comments here. When we think about the LTV differential between internal -- policies sold by internal agents versus external, what kind of magnitude you're talking about? Like how much is the LTV differential?

Derek, could you take that?

Yes. So what we have seen coming out last year is persistencies from internal agents written policies to be about 15% better than external agents. It's a good difference. The bigger difference, though, is the online enrollments versus telephonic enrollments. So last year, we saw online enrollments have higher LTVs, which correlates directly to persistency of 25% to 30% higher than telephonic enrollments. So our strategy had always been to increase online and digital engagement, and it's good that continue to focus on all that will all still help us in terms of churn and LTVs.

Okay. Let's talk about the current AEP. I know we're almost 1 month into the AEP. I was wondering if you can share any early takeaways in terms of how enrollment have been trending. I know you've talked about in the past that the last 2 weeks are very important to the AEP, most of the sign-ups happen then. But still like, I mean, any early read takeaways you can share from a enrollment perspective as well as from initial read from the impact of these initiatives you're putting in place?

Yes. AEP is off to a good start as we expected. We prepared well. And that's what Q3 is largely about for us is preparing for AEP. As I indicated, we increased our internal agent count by 120%. So we hit the ground running with the capacity that we needed. The election we knew would be a distraction. It always has been. My first year as CEO was 2016. And so the spot market for direct response TV unfolded as I expected. It actually ended up being even more expensive, upwards of 50% higher prices on the spot market. We had purchased almost all of our media spots in the summer at much lower rates. It was a combination of election, candidate spend but also proposition spending in a lot of states, not just in California, which is [indiscernible] anyone who lives here. But Illinois had progressive income tax on the ballot that was defeated there, and there were just hundred millions of dollars spent. And then -- so it wasn't just a presidential election. Some of the most expensive senate elections in history. I think the most expensive senate elections in history were this year. So we -- while it's always back-end loaded, the last 2 weeks, make or break it, but it will be more so this year. From our standpoint, having the higher agent capacity internal resource as well as having released our customer contact center and the significantly enhanced online enrollment experience with enhanced drug and provider databases, fuzzy logic capabilities, we are positioned to that. The more it's back-end loaded, the more we will relatively outperform because we always -- all of us, all participants in the telephonic channel end up under capacity the last week, right? The wait times average an hour. I mean it's just because seniors procrastinate. And they do so for reasons that are completely understandable. This is a highly consequential, stressful decision. They're afraid if they make a mistake that they're going to impair their coverage or their spouses' coverage because usually 1 person in a marriage relationship completes the applications for the other and -- or they are afraid that they will enroll in some plan that they're spending too much money. And so the combination of all of this just makes human nature's tendency to procrastinate exacerbated. And so this is why we think this year could be the inflection point for online, unassisted enrollment. So I'm very excited about what might happen, but don't read more into that than I can say because it hasn't happened yet.

That's helpful color. I mean, good to know that. But yes, we have heard about that as well. Actually, 1 e-mail question is coming in. Let me take that because it's very timely on a comment you just made. What explains lower churn from online enrollment versus enrollment that involves an agent? Some competitors have talked about having the opposite experience, like selling higher churn with online enrollees. I mean seeing higher churn there that don't interact with agent. I know you have talked about that. Do you want to spend some time on that?

Yes, for sure. And it's important enough that I'd like you to hear Derek's color on it as well, whatever I don't say. And not all of this is known by us. We, of course, are doing user research and survey, and we have a whole team now that -- someone that Derek recruited from his past life is heading up, and she's doing a fantastic job. But our understanding of seniors this year compared to last year, it is really night and day, but we're still the early days. We're just now scaling in terms of the online, unassisted. But I'm telling you, it is -- it's indisputable that eHealth's online enrollments have significantly higher persistency. The most important thing -- strip all the noise away around churn and LTVs, the most important thing here is getting the senior in the right plan. When we get a senior in the right plan, we have better consistency -- better persistency rather. And the challenge that we've had, of course, is the databases are not one uniform database that you can go collect all the drug information. There's not one uniform database to collect all the provider information. And so even -- we have -- before our upgrades, we have the, by far, most complete database and accuracy, but still way too many holes in that drug database and the provider database. So with that big improvement, we're expecting another leg forward of improvement. So we think that's what differentiates us. Now we spent tens of millions of dollars on this. And so it ought to be better, right? And so it's not surprising that the experience that our competitors are having is different than ours. But the other thing I should point out that the agency that a senior has when they're enrolling, by -- almost by definition, they're more knowledgeable about Medicare. They've shopped, they've made a considered decision, they've entered their drugs, they've entered their provider information, they've entered their preferred pharmacy. That takes the level of sophistication that the average telephonic responder to a direct response TV ad doesn't possess, okay? So they tend to be younger. So we have a debate within the company. Some believe that this has nothing to do with the experience. It just has to do with who's doing it. We have one person who says, "Oh, the only reason people are staying longer, it's so damn difficult to enroll in our site prior to our enhancements that they just -- the brain -- they have PTSD from doing it." So they stay on for a few years before they want to consider it again. I hope that's not the reason. But we've made it a lot easier to enroll. But now with this customer center, where we have 2-factor authentication for the registration. And the seniors are entering their drug information, they're entering all of their provider information, we think the stickiness of those enrollees could be well in excess of -- well, that was already very attractive persistency. Derek, what's your view?

Much the same, the sophistication of the people who enroll with us online is a key driver, so -- which then translates into they're more likely to get into plan that works for them because they've entered right information, they understand the plans, and they understand use of the plans and so on. The agency piece, just a nuance to kind of expand on what Scott said is we do, from user research, find that people who take ownership of making the decision more likely to stick with it and, therefore, not respond to the next ad that they see this that say you should go look at your Medicare Advantage insurance again. And it makes sense. When you're doing it at least partially even online, you feel like that you're more accountable as opposed to being sold to, right? We all take more accountability and we trust our own judgment more than what someone else want me to do. So that's the reason we see the online persistency to be higher.

Okay. That's a fair point. I want to follow up on one item, which actually got -- caught a lot of attention coming out of third quarter earnings call was the sequential slowdown in approved MA members growth. You had growth in the quarter at 45,000 3Q '20 versus a little over 60,000 in the second quarter. I know you talked about that second quarter did not have much SEP benefit. So what explains this decline of 15,000 sequentially? And are you still comfortable with 70% plus growth in the fourth quarter?

Yes. So we did have some benefit in Q2. It's hard to completely unpack that. Again, these are the kinds of things. There's some differences of opinion within the company as how much impact SEP had in Q2. Q3 is an unprofitable quarter for us. And I think it's important to understand that the enrollments -- the cost per enrollment is high. And so one wants to be conservative about how much of our full year spending we allocate to Q3. We still need a critical mass at it because we need live calls to train that 120% of increased agent capacity that we've hired. So it's a fine line. We need enough to train so that when -- by the time we get into the heavy demand period that we have a well-trained internal agent sales force. And I'll make an even finer point. The profitability of direct response TV that runs in the first half of AEP is much less than the profitability of DRTV that runs in the second half of AEP, okay? That's our biggest, biggest demand period. So the profitability of direct response TV in Q3, you can imagine, is negative, right? And so we're not highly incentivized to aggressively drive demand. So what was really positive about Q3, and I didn't do a good job of explaining it, was we had the discipline of not spending a ton of money on those direct channels like response TV. And we got a very large percentage and over 1/3 of our enrollments were online. And so that's the more profitable enrollment, and that's going to have the longer persistency. So the mix of Q3, the ultimate profitability mix of Q3 when their true LTVs work their way through our 606 regression analysis, it's going to be very attractive. And I'm quite confident, we're beyond hopeful that the same will be true for Q4. We -- and in the number is about 60%, Jailendra, for the MA growth that our midpoint of our guidance contemplates, look, that's down from 100% last year. But like I said, remember, too, that there were a couple of deciles of enrollment last year that weren't profitable for us. So you start with -- we're not doing that again. A lot of effort for no profitability, in fact, negative profitability on the full analysis. So we -- what we're maintaining the discipline, we want to grow. We want to capture market share. We want to continue to be super important as a partner to our carriers. But at the same time, we are highly committed to improving the quality and the profitability of our enrollments. And so that's why we -- for us, this is a really modest growth rate. I know that it may seem aggressive, but the growth is out there. The opportunity to grow is there. We just want to be sure that we have the discipline, even if it takes a slight reset, which is what we regard that midpoint of the guidance to be this year, we want to over-index on profitability even if we underperform in terms of share with our public company comps.

That's a fair point.

Yes. Let me just add one more thing is when we reversed guidance in July, our internal plans was have -- was to have Q4 to be a disproportionately higher growth quarter than Q3. So what we saw in Q3 was within our expectations in terms of quarterly sequential growth. I shared with Scott, that we could have done a better job in articulating that when we kind of published our guidance. Clearly, there is not a consensus on that view on what people expected. So that's something that we'll do better on. But what is important for people to know is, to an extent that we are managing allocation of investments and managing risk, it is always better for us to drive more enrollments in Q4. And to Scott's point, if we can do, we'll drive it all on the last day. But of course, that's theoretically impossible, unless, there's a lot of enrollments, that is theoretically possible. So we -- that's what we want to do. Again, we'll take the learning from the fact that we could have done better in terms of just letting people how they played out in the back half of the year between Q3 and Q4.

Great. So I mean, last year, I think Scott mentioned this that you guys got a better sense of churn rate when you reported Q1, right? I think you were -- things were higher than your initial expectations. One pushback from investors we get is like are we going to wait until, again, Q1 '21 earnings that we won't have good visibility whether all these initiatives are really yielding results or not? How can you make investors feel comfortable? I mean, but we still have almost 5 months to go by the time you report Q1. So can you provide any color on your Q4 call or anything over the next 5 months, which could be a good leading indicator that you are moving in right direction?

Well, I never know what Kate will permit me to say -- so that -- in the future. I know what she's permitted me to say today. But we will know -- I'll be curious what -- I will know on December or some date, pretty close to what our persistency will be from our enrollments because of the fulfillment source. We will know what percentage have come from online lead gen and then what percentage were fulfilled online assisted and unassisted? How reliant we ended up having to be in direct response TV? If we had to step on the accelerator, and that's what we did last year just because we saw that we could, we stepped on that accelerator and then regretted it. So we will not do that this year, okay? So we're not going to -- we won't make that mistake. But the less reliant we are on TV, the more profitability, the longer persistency we will have. And so that's the toggle for us. And if -- I've been asked this, if push came to sharp, we've got the 60 -- we have to hit 60% in order to get to the midpoint of our revenue guidance. And remember, the LTVs and the revenue guidance has a 10% decrement here. Almost $35 million will be just arbitrarily taken off of the top line because of the look -- 8-quarter look back. So if -- again, it's real and it's the correct accounting. I'm not saying that it's not appropriate, it's correct. But it bears no relationship on what the actual value of those enrollments will be. But by maintaining this discipline, not growing over 60%, okay? By not doing that, it gives me the assurance that we will bring in high-quality, long persistency enrollments. So that's the key here. And we won't -- my intention is to not go over that unless it happens without my ability to turn off the spigot, if it just -- because if it comes through online. And if we end up overperforming, I don't want you to call me a sandbagger. It will just be because we get an avalanche of last week -- last few days online enrollments because that's infinitely scalable. And so it could happen. But if it does, it should be very high-quality overperformance.

So just to make sure I understand. So in Q4, is there a possibility of upside to your LTV trends, which we expect to be down 10%?

No. It's baked in...

Do you think that's kind of done, that's written in stone.

It's baked in. It's my reward for driving the growth I did last year without attention to LTV, all right? And so we've acknowledged it and said that it was, with the benefit of hindsight, a mistake. And we've remediated that in every aspect of our business processes. But we -- so we don't have upside to reported LTVs per enrollment. They are dictated by our regression analysis, and it's correct. But when you think about our profitability for next year and when you look at our guide, and it may seem aggressive compared to this year. But based upon our assumption that we get back to '19 levels, that's a lot of deadweight that we got -- that we carry this year that we get relieved of next year. And so that's why you'll see a very substantial uptick in adjusted EBITDA when we guide for 2021 against 2020.

Okay. The last one before wrap up, I want to touch upon the acquisition cost, like customer acquisition cost, you guys talked about down 10% for the year, kind of implies like down 13% for the quarter. How is that split between trends on CC&E costs versus marketing and acquisition cost and the key drivers, can you highlight those?

Yes. I'll go real fast and then let Derek, yes, be more specific. One area that I'd like to highlight that's a driver here is the partnership channel, which will exceed 20% of our revenues this year. And is on a nice trajectory. We have lower cost of acquisition. Most of those partnerships like the provider channel, hospitals, pharmacies, they cannot even take a rev share. So they partner with us and their costs associated for us in building those partnerships and marketing to those partnerships, but they are lower cost of customer acquisition. They have higher conversion and they have longer persistency. And we've been investing heavily in '18 and '19 on this channel. We were in a net investment status up until this year. We are just crossing profitability this year in the partner channel, and then we should get meaningful profit contribution from the partner channel next year. Then the online channel brings down the CC&E, the customer care and enrollment expense,significantly because we don't have agent costs on the online unassisted. Derek, what would you add?

Not much more other than that we expect that the customer care enrollment item to have a disproportionately greater decline than the marketing, but we do expect both on a 4-year basis to decline year-over-year.

All right. I guess, we are out of time here. So we will leave it there. Thanks a lot for participating in our conference. Have a nice rest of the afternoon. Thank you. Take care.

Thank you.

Thank you.