Logo of jester cap with thought bubble.

Image source: The Motley Fool.

eHealth (NASDAQ:EHTH)
Q3 2019 Earnings Call
Oct 24, 2019, 5:00 p.m. ET

Contents:

  • Prepared Remarks
  • Questions and Answers
  • Call Participants

Prepared Remarks:


Operator

Ladies and gentlemen, thank you for standing by and welcome to the Q3 2019 eHealth, Inc. earnings conference call. [Operator instructions] I would now like to hand the conference over to your speaker, Kate Sidorovich. Please go ahead, ma'am.

Kate Sidorovich

Thank you. Good afternoon and thank you all for joining us today either by phone or by webcast for our discussion about eHealth, Inc.'s third-quarter 2019 financial results. On the call this afternoon, we will have Scott Flanders, eHealth's chief executive officer; and Derek Yung, chief financial officer. After management completes its remarks, we will open the line for questions.

As a reminder, today's conference call is being recorded and webcast from the IR section of our website. A replay of the call will be available on our website following the call. We will be making forward-looking statements on this call which includes statements regarding future investments, beliefs and expectations, including statements relating to our expectations regarding our Medicare business, including Medicare enrollment growth, consumer demand, our competitive advantage, our market share, growth in online enrollments, and our performance this annual enrollment period; our investments in the Medicare business, including investments in telesales capacity, online demand generation channels, sales and marking and our technology platform; our expectations regarding the profitability of our business, seasonality, lifetime values, policy durations, retention rates, conversion rate and cost of acquisition; our views regarding the Medicare environment and current political environment and our outlook for the fourth quarter and full-year 2019 guidance, including our assumptions and our ability to deliver on our guidance. Forward-looking statements on this call represent eHealth views as of today.

You should not rely on these statements as representing our views in the future. We undertake no obligation or duty to update information contained in forward-looking statements, whether as a result of new information, future events or otherwise. Forward-looking statements are subject to risks and uncertainties that could cause actual results to differ materially from those projected in our forward-looking statements we describe these and other risks and uncertainties in our annual report on form 10-K, and quarterly reports on form 10-Q filed with the Securities and Exchange Commission, which you may access through the SEC website or from the investor relations section of our website. We will be presenting certain financial measures on this call that are considered non-GAAP on the SEC Regulation G.

For a reconciliation of each non-GAAP financial measure to the most directly comparable GAAP financial measure, please refer to the information included in our earnings press release issued today and in our SEC filings, which can be found in the About Us section of our corporate website under the heading investor relations. And at this point, I will turn the call over to Scott Flanders.

Scott Flanders

Thank you, Kate, and welcome, everyone. Strong momentum in our business continued with another quarter of meaningful outperformance against our expectations. Third-quarter revenue of $69.9 million grew 72% year over year, driven by a 70% increase in approved Medicare members, and tail revenue of $11.5 million that we recognized during the quarter, which represents cash collections on plans beyond what we originally recognized as revenue under our conservative revenue recognition practices. Our adjusted EBITDA was negative $18.8 million, reflecting significant investment made in our Medicare telesales capacity, as well as marketing investments during the quarter ahead of the annual enrollment period.

GAAP net loss was $11 million. Derek will cover our third quarter in greater detail later on the call. During the quarter, we completed our preparation for the annual enrollment period. We expect to set another record for the company this AEP for Medicare enrollment volumes and for commission revenue.

We successfully expanded our telesales capacity, with total sales agent head count more than doubling compared to last year's AEP, including the successful launch of our new sales center in Indianapolis. Over the past few months, we tested and deployed new agent facing technologies that we expect to further enhance our customers' experience and increase agent conversion rates. These tools are now available across our agent base, including both our in-house and flexible agent force. Our marketing programs, which have delivered strong results over the past four quarters, have been further refined ahead of the AEP, with a greater emphasis on data and analytics, designed to supply our agents with high quality leads at attractive cost of acquisition.

Our website experience for seniors who choose to enroll online or via a mobile device has been enhanced significantly compared to a year ago. Combined with stronger emphasis on online demand generation channels, we expect these improvements to translate into meaningful growth in online Medicare enrollment volumes compared to a year ago. The improvements we have made have been implemented against the backdrop of a strong market environment. Based on CMS forecasts, Medicare Advantage enrollments are expected to grow 10% in 2020, the highest annual growth rate in more than a decade.

Increased affordability of MA plans, broader plan choice and greater plan complexity with new supplemental benefits being introduced by carriers make eHealth's broad selection of plans and carriers combined with our market leading online comparison platform more relevant than ever to consumers. Simply put, we have entered this AEP from a position of strength, allowing us to guide up to the high end of our 2019 revenue and EBITDA forecast at a recent investor conference on October 4. Our confidence in the trajectory of our business remains high. Turning to third-quarter performance in our Medicare business.

Total revenue grew 75%, and commission revenue grew 78% compared to a year ago. Sponsorship revenue from carriers grew 83% as insurers have increasingly recognized the advantages of eHealth's customer engagement platform, which has delivered enrollment growth well in excess of the overall Medicare market. We continue to rely on the diversified portfolio approach to our demand generation strategy. And within our marketing portfolio, we are increasingly emphasizing digital advertising initiatives, which target consumers who are more likely to enroll online with limited or no call center interaction.

To this end, 21% of our major medical applications in the third quarter were submitted online, compared with just 9% in the third quarter a year ago. Third-quarter performance in this regard represented four times increase in the number of online applications year over year. We believe we are delivering solidly on our strategy to significantly expand the online portion of eHealth's enrollment activity. Our individual and family Plan major medical business, which returned to growth in the second quarter for the first time in several years, continued to perform strongly driven by higher new enrollment volumes and increased member retention.

Approved IFP members grew 76%, and IFP commission revenue grew 95% compared to the third quarter of 2018. Turning to the small business market. The number of submitted applications declined 15% compared to the third quarter of 2018 as we shifted a portion of our sales resources to Medicare ahead of the AEP. small business group commission revenue increased 14% over the same time period due to higher renewal revenue.

In conclusion, this year we have delivered three consecutive quarters of successful execution, exceeding our expectations in every single quarter and organically growing our year-to-date revenue 76% compared to the same period a year ago. Annual enrollment period began last week on October 15, and we entered this important Medicare selling season from a position of strength and confidence. During this AEP, the majority of Medicare recipients have access to a broader selection of plans compared to a year ago, and our platform is an optimal destination for seniors to make an informed choice and enroll in a plan that best suits their individual needs. There's been a lot of activity on the political front ahead of the election year, but we believe that regardless of the outcome of the primary and the presidential election, the Medicare Advantage program will remain a cornerstone of administering health insurance to seniors in this country.

eHealth's marketplace focus, technology platform and consumer-centric business model is proving to be the best solution to help more and more seniors access the best health insurance for their individual needs. And now I will turn the call over to Derek to speak more specifically about the numbers.

Derek Yung

Thank you, Scott, and good afternoon, everyone. Our third-quarter results reflect strong revenue and enrollment growth in our Medicare and individual and family businesses and a significant investment in our telesales capacity and marketing initiatives ahead of the Medicare annual enrollment period. In our Medicare business, third-quarter revenue of $57.2 million grew 75% compared to a year ago due to a 70% year-over-year increase in approved Medicare members, 64% growth in noncommission revenue driven primarily by carrier sponsorship revenue and $3.8 million in tail revenue driven by greater than estimated customer policy duration, which I will describe in greater detail shortly. The Medicare segment generated a loss of $11 million, reflecting investments as we have prepared for what is expected to be another record fourth quarter selling season in terms of revenue, enrollment volumes and profitability.

Year to date, our Medicare segment profit was $5.9 million, compared to $2.2 million for the same period last year. Our estimated number of revenue generating Medicare members was approximately $551,000 at the end of the third quarter, up from promptly $410,000 at the end of the third quarter of 2018, or an increase of 34%. Third-quarter 2019 revenue from our individual, family and small business segment was $12.7 million, a 59% increase compared to a year ago. Similar to the second quarter, we saw a strong increase in the number of approved members for major medical IFP products, which grew 76% year over year.

In addition, we are seeing a continued trend of longer duration for these products, resulting in higher estimated lifetime values. Third-quarter constrained LTDs for individual and family Plan products grew in excess of 40% compared to the third quarter a year ago as a result of favorable commission rates and policy duration. Commission revenue in our IFP business grew 95% compared to a year ago. Small business group commission revenue increased by 14% year over year.

The individual, family and small business segment profit was $3.8 million compared to a loss of $0.6 million in the third quarter of 2018. Our estimated individual and family Plan membership at the end of the third quarter was approximately $131,000, down 19% compared to the estimated membership of $161,000 we reported at the end of third quarter a year ago. The estimated number of members on small business products was approximately $45,000, a 17% increase compared to a year ago. Our total revenue for the third quarter was $69.9 million, an increase of 72% compared to the third quarter of 2018.

Our total estimated membership at the end of the quarter for all products combined was approximately 991,000 members, including approximately 264,000 estimated members on ancillary products. Before I move on to discuss operating expenses, I want to provide more detail on the dynamics that we're seeing with estimated lifetime values, or LTVs, in our Medicare business. Under ASC 606 accounting, we recognize revenue based on the constrained lifetime value of each enrollment that is approved by our carrier partners. The base LTVs are determined through an estimation of commission payments that we expect to collect over the life of the approved policy business, and is based on a detailed historical analysis of several input, including our past experience with policy durations and commission rates per paying member.

These estimates of the base LTV are then reduced by applying a constraint factor, effectively discounting the revenue we both compare to what we expect to collect. This constraint acts as insurance against overstating revenue for the duration of the policy due to fluctuations in policy duration, commission rates or other external factors. Our revenue recognition approach using constrained LTVs is appropriately conservative, and this is demonstrated in how we recognize tail revenue in every quarter of 2019. We recognize tail revenue when our cash collections are in excess of the estimated constrained LTVs for previously sold Medicare policies.

The constrained LTVs for Medicare Advantage plans, which account for the majority of our existing Medicare members, have grown on a year-over-year basis for five executive quarters, driven by higher commission rates and better than expected policy duration on many of our existing policies. On our last earnings call in July, we noted that after several quarters of continuous growth in constrained LTVs for our Medicare Advantage products, we expected the LTV to decline by mid-single digits for approved applications in the fourth quarter of 2019 compared to the fourth quarter of 2018. This forecast was driven by higher policy turnover that we observed during the first quarter of this year on Medicare Advantage members that we enrolled during the last annual enrollment period in the fourth quarter of 2018. We believe that the increased turnover in that specific group of approved applications was driven by the open enrollment period, which took place in the first quarter of this year for the first time since 2011.

Since our earnings call in July, we have observed a reversal of this trend, with lower turnover on this Medicare Advantage cohort. As a result, the difference between cumulative turnover on this Medicare Advantage cohort on a year-to-date basis compared to how prior fourth-quarter cohorts turnover by this time has narrowed. At this point, and in light of these circumstances, we expect that fourth-quarter Medicare Advantage constrained LTVs will only be down by 1 to 2 percentage points compared to the Medicare Advantage constrained LTVs for the fourth quarter of 2018. For the full-year 2019, our weighted average Medicare Advantage LTVs are expected to be flat to slightly up compared to 2018.

Our commission receivable balance at the end of third quarter was $358 million, an increase of $16 million compared to the ending balance at the end of the second quarter, and an increase of $92 million compared to the balance at the end of the third quarter a year ago. Our commission receivable balance reflects future commissions we expect to collect on our existing membership, discounted by the constraint factor that I described earlier. Our commission receivable balances increases each quarter by the amount that we recognize our revenue for in-quarter enrollments, which is based on lifetime value estimates. This increase is net of the upfront commission payments for in-quarter enrollments and reoccurring commission payments that we received on our existing members enrolled in prior periods.

During the third quarter, we collected $43.4 million on commission payments. Now I would like to review our operating expenses and profitability metrics. In the third quarter, we made significant investments in expanding our Medicare telesales capacity, including our new Indianapolis sales center launch as well as hiring, training and onboarding agents ahead of the October 15 start to AEP. As Scott mentioned earlier in the call, we grew our agent count in excess of 100% compared to the same time a year ago through a combination of in-house hiring and expanding our outsource agent force.

Our third-quarter non-GAAP customer care and enrollment expenses, which excludes stock-based compensation, grew $22.7 million, or 133% compared to the third quarter a year ago, reflecting this investment as well as the fact that we are operating with a larger agent count this year, having kept the majority of our in-house agents on board following the completion of the last AEP. The newly hired agents were not fully productive for most of the third quarter, as they were going through licensing and training. As a result, customer care and enrollment expense per approved member increased significantly both sequentially and on a year-over-year basis. Similar to last year, we expect this metric to come down in the fourth quarter when our expanded telesales resources are leveraged across a much larger volume of enrollment and as the agents become more productive.

Our non-GAAP marketing and advertising expense, which excludes stock-based compensation, grew $9.3 million, or 60% year over year. Variable marketing expense in our Medicare business grew 78% compared to a 70% growth in approved Medicare members as we invested more on the margin in direct online demand generation channels that have higher marketing costs of acquisition, but also tend to have higher unassisted online conversions. Shifting toward increased online penetration, which should over time translate into higher operation scalability and lower agent cost per enrollment is at the core of our growth strategy in the Medicare market. Adjusted EBITDA for the third quarter of 2019 was negative $18.8 million, which was better than our expectations and compared to negative $6.9 million for the third quarter of 2018.

We calculate adjusted EBITDA by adding restructuring charges, acquisition costs, stock-based compensation, change in fair value of earning liability, depreciation and amortization, amortization of acquiring tangibles, other income and benefit from income taxes to our GAAP net loss. GAAP net loss for the third quarter of 2019 was $11 million, compared to GAAP net loss of $9 million for the third quarter of 2018. Our third-quarter cash flow from operations was negative $15.9, compared to negative $5 million for the third quarter of 2018. Capital expenditures, which include capitalized internal developed software costs were approximately $4 million for the third quarter.

Our cash balance was $91.4 million as of September 30, and we had no debt outstanding under our line of credit. Based on information available as of October 21, 2019, eHealth is reaffirming its guidance for total revenue and adjusted EBITDA for the full year ending December 31, 2019, and also updating its guidance for certain GAAP and non-GAAP measures for the full year ending December 31, 2019. Due to the third quarter of 2019 reduction in fair value of the earning liability assume in connection with our acquisition of GoMedigap and the change in income tax rate. Our 2019 guidance ranges are included in our third-quarter earnings release for your reference.

Based on the quality and scale of our call center resources in place, the acceleration of our online enrollments guidance ranges and the strength of our consumer demand we're observing, we are confident in our ability to deliver at the high end of the revenue and EBITDA guidance ranges. This implies a year-over-year growth in 2019 approved Medicare members of over 60%. After what has been an investment quarter, we are forecasting a return to meaningful EBITDA profitability in the fourth quarter, driven by significant sequential increase in enrollment growth and a related increase in our revenue. On a year-over-year basis, we forecast our fourth-quarter EBITDA to grow significantly both on a margin and absolute-dollar basis relative to the fourth quarter of 2018.

The top end of our 2019 EBITDA guidance implies fourth-quarter 2019 EBITDA growth of over 50% compared to Q4 of 2018. I want to remind you that these comments in our guidance are based on current indications of our business and our current estimates, assumptions and judgments, which may change at any time. Our actual results may differ as a result of changes in our estimates, assumptions and judgments. We undertake no obligation to update our comments or our guidance.

Before we turn the call over to the operator, I want to address our membership metrics in context of customer policy retention, a topic that's been top of mind of our investors and analysts over the past few months. In order to compute customer policy retention rates, it's important to keep in mind the following three dynamics: one, our retention rates are more accurately assessed by using the number of paying members that we add in a given period, and not the number of in-period approved members. Our approved Medicare Advantage members have converted into paying members at an average of 92%. In addition, some of the members that are approved in a given quarter don't become paying members until the following quarter, a spillover effect that's especially pronounced during high enrollment quarters.

Second, policy turnover is higher in the first year post approval and starts to decline in the outer years. Policy persistency is particularly strong once a policy has been held for over a year, getting us to our current average policy duration of approximately three years for Medicare Advantage. Three and lastly, due to the pronounced seasonality of our enrollments, it is more accurate to look at churn and other metrics on trailing 12-months basis. We summarize these points and also provide an example of customer policy retention math as part of our earnings call slides, which are now posted on our customer relations website.

Our goal has always been to be fully transparent to our investors and analysts, and over the coming quarters we will be revisiting our reported metrics and to evaluate if we can make additions or changes to further that goal. And now we will open up the call for questions. Operator?

Questions & Answers:


Operator

[Operator instructions] Our first question comes from Jailendra Singh with Credit Suisse. Your line is now open.

Daniel Aversano

Hi. This is Daniel Aversano filling in for Jailendra Singh. Thank you for taking my question. Do you have any more details on the new call center in Indianapolis? I know last year you had approximately 136,000 unanswered calls.

So have you seen less leakage so far? And lastly, can you talk more about your marketing investments and engagement efforts?

Dave Francis

Yes. Hi, thanks for the questions. This is Dave Francis. The Indianapolis call center has launched, really, without a hitch.

I think we talked about previously that we were looking to staff as many as 200 sales people within that office, and it's gone, like I said, without a hitch. We had full hiring, had everybody trained, appointed, and when the AEP bell opened on October 15, that was our first office opened and taking calls, and they've been performing as or better than expected. As it relates to the number of calls that we've talked about last year that we were unable to answer because of constrained capacity, so far we're early in AEP, so it's way too early to be telling exactly what the full AEP will be, but we feel very good about operationally how all the call centers are operating and the number of customers that we were able to touch. We've got meaningfully less leakage, as you put it, than we did last year.

And as it relates to the marketing side of things, everything is working as we expected. We've made some significant investments across the different marketing channels that we talk at length about, all the way from direct mail up to search engine marketing, and very pleased right now with the way AEP is going. With that said, we're less than 20% of the way through AEP, so it's way too early to see how everything's going to end, but we feel very good about where things stand at the moment. And I think that's all we're prepared to say at the moment.

Scott Flanders

The only thing I'll add to that is in Q3, evidence to our ability to serve more calls is that we did have less call selling revenue in Q3 this year than compared to last year given the productivity and the size of the telesales force that we had this year compared to last year in Q3.

Daniel Aversano

OK. Great. Thank yhou. And just one more quick question.

Would you also be able to talk about some of the initiatives you're focused on for your lead sourcing and bringing down your lead sourcing and customer acquisition costs? Do you see a big opportunity there?

Derek Yung

Yeah. I can start with that. We continue to have success in scaling our direct to consumer marketing and customer acquisition. It's been a multiyear effort.

And for many years, as you have seen, we have driven the cost per member down on marketing. Our focus now is really continue to keep that scale while growing enrollment. So just to kind of echo that point, our desire here is not to look for a reduction in new economics of marketing, it's to maintain it while we scale. And a key component to that is continue evolution of our marketing program to be more analytically driven, which was discussed by Scott, and the other highlight here is we will continue to expand and increase our online marketing at a more outsize rate compared to other channels.

And the reason for it is that the members that we acquire through online marketing channels tend to enroll more online.

Daniel Aversano

Perfect. Thank you very much.

Operator

Thank you. And our next question comes from the line of George Hill with Deutsche Bank. Your line is now open.

George Hill

Hey. Good afternoon, guys. And I hopped on the call a little bit late. I'll apologize if you cover this.

I guess, Derek, you talked about 90 -- I'm trying to read my notes here, approved members becoming paying members 92% of the time. I guess, can we talk about the other 8%? And can you frame this statistic relative to history, or is this new?

Derek Yung

George, that's a good question. So and I'll point people to what I commented in my prepared remarks, which is more details on the slide that is we posted in our investor relations website. The 92% have been very steady in the history that we've been in this Medicare business, and even in the figures that we have disclosed on the slide. You can see that it averages out to 92% when you look at on a trailing 12-month basis the last four quarters.

What tends to happen is people either don't actually qualify and end up not paying, or they end up not paying because they are unhappy with the plan, and it's not a unusual situation not just for eHealth, but for other brokers or other plan providers where they see a drop of return approved pay.

George Hill

OK. And like I said, I apologize I dialed in a little bit late. It sounded like you said for Q4, the erosion in LTVs is now going to be 1% to 2% versus mid-single digits. I guess there's a couple of factors that drive the LTV component.

I guess can you talk about which puts and takes have moved versus when we communicated at the end of Q2? My guess is going to be that retention is looking better because you have pretty good visibility of the other components of the LTV calculations.

Derek Yung

Yeah, that's good insight there, George. So the three primary inputs in LTV is policy duration, commission rates and the approve to paid ratio that we had just commented on. The latter two of those components are fairly consistent. One's obviously a cohort has enrolled, so it really is around the persistency of the policies.

As I mentioned in the prepared remarks, which I get more color here, what we saw was coming out of the OEP period, people who have enrolled during AEP in 2018 churning more, meaning they switched more because of the ability to be able to do that given that OEP period. Since then, for people who did not switch, what we've seen is a reversal of that trend where they're staying on longer than what we've seen compared to historical patterns. So when those two things come together on a cumulative basis, we're seeing the churn basically narrowing. So it's not obviously the full-year cycle yet, so the final chapter of this is still to be seen, obviously, with AEP.

But at least we can say, given the evidence we've seen so far, we are seeing perhaps a more of a shifting in the seasonality of people shifting as opposed to maybe a permanent increase in switching based on the introduction of -- or the reintroduction of OEP.

George Hill

All right. That's helpful. And I don't want to neglect Scott and Dave. So what I want to ask about is I want to know how are you guys thinking about the Q1 switching season.

I guess how do you retain applicants versus people who came in in Q4? You probably can't do anything to prevent switching, but I guess what I care about is from a marketing, go to market perspective, do you guys keep the pedal to the metal on marketing spend and the attraction of beneficiaries in early applicants in Q1 given this is the second year that we have the switching season, and whether it's people aging in or people who are switching? It's almost like you get a little bit of a second bite of the apple as it relates to AEP. I'd love to know how you guys are thinking about the switching period kind of second time around.

Dave Francis

So george, it's Dave. I'll go first and let Scott pile on. But there's two things as we look at Q1. Number one is, for lack of a better term, playing defense on those comfort that is we just signed up.

To the extent that we can limit that increased turnover in customers that we saw in the first quarter of this last year from the previous Q4, new buying cohort, we've got some plans in place to have specific outreach for those folks to make sure that they understand what they bought, if they're using it properly, if they have any questions, to help those folks better understand and make sure that they're correctly using their product so that there's less impetus to want to turn over. If they do want to turn over, then we're right there for them to do that with us so that we keep them in our book and they don't end up exiting the book. Then secondly, as we've talked about and have demonstrated through our outsized growth rate relative to the market, any time of opportunity for broad-based customer groups to change, we believe they ought to be doing so using our platform given the tools and the breadth of choice that they have. And as we saw last year with significant growth in the first quarter, we fully expect to be plowing additional marketing resources into the marketplace so that we can capture as many of those customers as possible.

So it's increasing the persistency of all of the customers in our book, including the new ones that we've added in the fourth quarter, plus encouraging those folks that are coming back into the market in the first-quarter OEP to come to the eHealth platform to do their shopping.

Scott Flanders

What I would add to that is those members that churned out from AEP in last year's fourth quarter and in new plans in Q1 have shown low churn, and so what is probably the case is they got into the nonoptimal plan, and then got into the right plan in Q1 and have had strong persistency since then. And so this is why we flagged the higher churn, and Derek might have been lost in a very long CFO script because we covered a lot, but we've been pleased to see that that churn abated dramatically. And so we did flag it when we thought it was a potential issue, and from today's call we're trying to alleviate anyone's ongoing concern about churn for us. It's flat year-over-year overall.

Operator

Thank you. And our next question comes from the Tobey Sommer with SunTrust. Your line is now open.

Jasper Bibb

Thanks. This is Jasper Bibb on for Tobey today. You have invested a lot in technology in the online enrollment experience. It appears others are likely to follow.

How long would you say it would take for another platform company to get to where you are right now?

Dave Francis

It's Dave. I'll answer it this way. We've talked a lot about our platform, both from a telephonic and an online perspective. There are two aspects to the platform that are important.

One is the front end experience for the customer where they engage with us. And candidly, we've seen a lot of people coming in and creating online quoting and shopping experiences where they're able to show different customers' plans and that sort of thing. We would call that the nonproprietary side of the marketplace business. We continue to invest heavily on the front end of the shop and make sure that our customers have the best shopping experience available to them.

The back end of the experience is what we believe is proprietary to the way that we're doing business that we connect across all of our products with over 170 insurance companies across the country, large, national, small, regional and otherwise. And that is the part of the business that is very difficult to replicate given the inconsistencies of technology platforms across all the different payers, and creating the connections and maintaining all the business logic and what have you that's entailed in all of that. That's a gross oversimplification of what it takes to build that back end, but it's something that we believe take not quarters or months, but years for another potential competitor to replicate if the insurers were willing to open up their back ends to that kind of electronic linkage. So we see a lot of people coming into our marketplace given the growth that we've had on the front end, the customer engagement side.

And we believe the more folks coming in, the better. It opens up more shopping experiences for customers generally. But to put the full marketplace end to end experience together in a customer-friendly way that we have, we believe, is extraordinarily difficult, expensive and time-consuming for a potential competitor to do.

Jasper Bibb

Thanks. That's very helpful.

Operator

Thank you. And our next question comes from the line of George Sutton with Craig-Hallum. Your line is now open.

George Sutton

Thank you. You mentioned that tail revenues in the quarter were $11.5 million, and they've increased five quarters in a row. Curious, does your Q4 assumption include any tail revenues?

Derek Yung

So that's a good question. Just to be clear, we've seen tail revenue five quarters in a row. I don't think we've specifically commented that it's increase five quarters in a row, but $11 million for the quarter is correct. So in our guidance, there is expectations included around tail revenue, and that's actually been the case since our first guidance.

What is also true is tail revenue have exceeded our expectations, and relative to the earlier questions from George, really when we see better than expected results in our cash collections, it is driven mostly because of better persistency, people staying in policies longer, so that's the case. So obviously, we have analytics around a fairly large membership of people. The longer that someone's been on a policy, the less variability there is around churn. And you actually can see in some of this data that we also have included around the life cycle of churn percentages, at least for Medicare Advantage, in our investor relations materials.

So more to come on that, obviously, based on how we see Q4 plays out. But hopefully that answers your question, George.

George Sutton

Wondered if you could explain your growth plans for the AEP versus the combination of your increased capacity and the assumption of higher online-only ads.

Derek Yung

George, what are -- is there -- are you talking about that...

George Sutton

So to be clear, you are assuming higher online-only ads, but your growth rates are obviously quite a bit different, so I just wanted to make sure I understood that dichotomy.

Scott Flanders

Look, in a word, I would just say that our guidance is conservative.

George Sutton

That was my assumption. Last question. Are you seeing -- or let me ask you it this way. Industry folks, including us, have looked at Medicare plan finder and found it to be relatively uncomfortable and unlikely for most people to work through.

Are you finding any impacts thus far in the season from Medicare plan finder?

Scott Flanders

No.

George Sutton

Thanks, guys.

Operator

Thank you. And our following questions comes from the line of Frank Morgan with RBC Capital Markets. Your line is now open.

Frank Morgan

Good afternoon. Just going back to the AEP. I'm just curious. Do you have any specific numbers you might be able to share, like how it's going so far? I mean, presumably you're getting realtime updates on those numbers and understanding that you're early in the process.

But any way you could quantify or give us some kind of benchmark about how it's going so far in this AEP?

Dave Francis

Frank, you're right. We see the numbers on an almost real-time basis. I would say that we are very pleased with the way that we planned for and -- planned for AEP and the way that the numbers are coming in thus far.

Scott Flanders

Yes. Record calls and record conversions.

Frank Morgan

OK. Thanks. And then just curious, a lot has been mentioned about the plan finder, potential for growing competition. But just curious.

With seeing strategic buyers in the space, does that in any way affect your view of how the competitive landscape plays out? And then my last one was that the 21% online number that you had in the third quarter, is that a fair number that you expect to see in the 4Q with AEP, or does that get diluted down somewhat given the growth in overall volume? Thanks.

Derek Yung

So let me answer the second question, is on a sequential basis, We have seen historically and we do expect fourth-quarter share of online applications to go to up. So we would expect the same thing to happen in 2019 as well. On the competitive side, and I'll start and then Dave and Scott can add, we play in a very large market as you know, Frank. And we've had a lot of success driving toward the growth strategy that we've been on the last four quarters.

And as you heard earlier in our prepared remarks and also in Q&A, we are very happy with and I'm very confident in our preparation and execution toward AEP so far.

Scott Flanders

Yeah, I'll leave it there. I do want to say, Frank, again, the fact that we got to over 20% online enrollments in the third quarter, as Derek said, the trend is for that to accelerate into the fourth quarter historically. And thus far this quarter, we have seen just that. We have high expectations for the transition to the online side of the business to continue to be a meaningful driver for us, both from a PNL perspective and a strategic positioning perspective, so we're really pleased with where those numbers are coming in, particularly given that we are only looking for 19% to 20% online for the entire year.

So we're really pleased with that. Again, competitively, we're pleased with what we're seeing in the marketplace relative to our ability to acquire customers.

Operator

Thank you. Our next question comes from David Styblo with Jefferies. Your line is currently open.

David Styblo -- Jefferies -- Analyst

Hi. Good afternoon. Thanks for the questions. Hey guys, the stock's down a few dollars in aftermarket trading.

I think a couple things that just are raising questions for investors that I thought I'd ask to see how you could help maybe add some clarity to it is, first of all, on the variable marketing costs for the approved members and customer care enrollment, those are obviously up 7% and 52%, as you guys released. How do we think about that normalizing? How quick does that maybe get back to flattish? Is that something that can happen in the fourth quarter since those agents will be seasoned, or does it take a longer period beyond that? That would be the first question I had.

Derek Yung

Yeah, good question. So the third quarter, both those metrics largely came in line or exceed our expectations given the preparation that we were looking to be put in place in terms of capacity going to AEP. So the way that we look at those metrics is really on back half of the year basis, and it will normalize largely because of, obviously, the AEP period, and also on the sales side, really getting agents to be productive, which many of them weren't in the third quarter because they were ramping in and getting training, getting licensed. On the marketing side, as we had commented in our prepared remarks, really the take up there was online marketing spend as a percentage of total marketing spend.

And because people that we acquired through online marketing channels have high propensity to go online, there's a great ability for us to up on marketing and have what we have left in the cost. And as you heard, we did exceed our expectations in Q3 for online enrollment, so you see a corresponding positive result from that effort. So yes, for the year, at the midpoint, we had guided to a lavish on the total variable cost. I think that's still the case.

We will obviously always look for additional investment opportunities to the extent they make since. And if it does make sense, it will more likely come from the marketing side through increase in online marketing investments.

Dave Francis

But to be very specific, Dave, the CC&E costs, the sales costs, are directly related to the fact that we need significant investments to put on significantly more sales capacity in the third quarter. The vast majority of that capacity was unproductive as they were going through onboarding, training and appointments. So we absorbed the cost in anticipation of the business coming in in the fourth quarter and having the capacity to help as men customers as possible in the fourth quarter, hence the third-quarter blip. And it's consistent with prior years as well.

David Styblo -- Jefferies -- Analyst

Sure. That makes sense. Thanks for the color there. I think the second element that could use some explaining is just on the LTVs.

The MA LTV was only up 1% this quarter versus the 8% to 16% in the first half of the year. I guess I would have thought it would have been up, again, pretty strongly just because you guys have a rate tailwind. And sounded like from some of the comments earlier that duration was looking better on the new cohort. So can you help me understand the discrepancy on the MA part? And then also on the MedSup LTV, the declines there accelerating up to 10% versus 3% to 6% in the first half.

Can you explain what's happening on the LTV on that side as well?

Derek Yung

Yeah, let me start with that. On the MS side, we acquired GoMedigap about a year and a half ago. And then we've really, really scaled that business since we acquired them. And we are going through a certain degree of growing pains around the operational aspects as we look to maintain or increase persistency, so you caught that.

On the MA side, there is a lot going on when we look at third-quarter LTVs. One is, obviously, there's a seasonality aspect of the LTVs. So if you look at it on a sequential basis, you will see some fluctuations. On a year-over-year basis, we use a two-year look back to average out the impact of either rate increases and also policy duration.

So that's why earlier in the year in early quarters, we had favorable longer durations that were still flowing into LTVs that were starting to get muted because of the enrollments we saw in Q4 2018. And so we talked about it in our prepared remarks. We are seeing that trend reversing, and normalizing and narrowing compared to what we saw earlier in the quarter, but it hasn't been completely reversed yet. So that's why you see more of a flattish behavior in Q3.

David Styblo -- Jefferies -- Analyst

OK. And then so that's not just a fourth-quarter event, because I thought initially the fact that the fourth quarter of '18 was sort of overestimated, you're correcting for that in the fourth quarter of '19, which the original guidance was down mid-single digits. Now it's only down one or two points. I thought that was the catch up, that it doesn't really happen in quarters 2 or 3 is -- but it sounds like it does actually happen throughout the year then.

Derek Yung

It does, but it's more dramatic for the quarter that it relates to. So our estimation's on a two-year look back, but is seasonalized, so that's the nuance. So just to comment more specifically, the churn that we were addressing both in the call in Q2 and now is specific to the AEP cohort. But in terms of LTV calculation, it does build into other quarters as well, but not nearly as much as the quarter that the season is related to.

David Styblo -- Jefferies -- Analyst

OK. Great. Thanks, guys.

Operator

Thank you. And our following question comes from the line of Greg Peters with Raymond James. Your line is open.

Greg Peters -- Raymond James -- Analyst

Good afternoon. I have one question and then a follow up. You mentioned competitive conditions. There seems to be a new player or a strategic coming into the market almost on a daily basis.

One of the things we've been hearing is rising lead costs. And I was wondering if you could give us some additional color around your go to market strategy and your process to drive the online enrollment. I noticed it was up 21% in the third quarter.

Derek Yung

So I think -- so if I understand your question is, have we seen additional competitive pressure against our marketing initiatives, and how has that shown up in terms of our acquisition costs?

Greg Peters -- Raymond James -- Analyst

Well, I'm sorry if it wasn't clear. We're hearing specifically that the cost of leads has been rising beyond just the 7% that you reported in the third quarter. And in the context of new players and/or strategics coming into the market, I'm just curious about your go-to-market strategy, how you're dealing with those pressures.

Derek Yung

Got you. OK. In Q3, the increase that we reported is primarily related to increase in online marketing investments and less so of competitive pressure broadly across other channels. And our principal channels are the direct channels with direct response TV, direct mail and direct traffic to our website.

So there was nothing in Q3 that we saw that would point to a broader increase in competition. With that said, you are right that there's a lot of interest in this market, and for good reasons because they probably see the same thing, obviously, that we see around the market opportunity, and obviously the growing aspect of Medicare Advantage as a very, very attractive product for consumers and also insurance carriers. So we're aware of that competition. We do feel like we have advantages in marketing that allow us to be able to compete in a manner that allows us to continue to scale, and most importantly as we do a lot of our lead acquisition principally through direct to consumer as opposed to relying on lead gen partners.

And then, of course, we continue to scale our online capabilities both from an enrollment perspective and marketing perspective.

Dave Francis

Greg, it's a competitive marketplace. But we continue to believe that we have the opportunity to grow at multiples of the market rate while still maintaining a very margin-positive impact on all the incremental business that we're bringing in despite the competition.

Greg Peters -- Raymond James -- Analyst

Right. My follow up question, which is -- and I understand the dynamics of the near-term growth opportunities in sacrificing cash flow, or free cash flow, that is. But I was wondering if you could just give us an update. I think in the past, you might have suggested that operating cash flow, or cash flow from operations, might turn neutral or slightly positive at some point a year or two down the road.

Maybe you can just give us an updated perspective on that.

Derek Yung

Yeah. At our investor day back in May, we had commented that we'd be close to break even on the trailing 12-month basis March 2020 based on the guidance at that point. We have obviously since revised our guidance upwards. And then we had commented in other settings that we don't expect to be able to be at operating cash flow break even at that same time period given our business model, which is heavier investments up front for customer acquisition.

So as of now, we expect that the cash on our balance sheet plus our access to a revolving facility to be able to allow for us to pursue our growth strategy for both in the five-year base scenario and also the tailwind scenario that we had discussed at analyst day. If there are additional opportunities to pursue more growth, we will look to leverage more on our $358 million commission receivable to give us more access to capital to review that growth. We will be revisiting our long-term growth plan, obviously, after this year since we are pacing ahead of both the base and the tailwind scenario in that five-year plan. And at that point, we'll be more able to discuss where we see our cash flow dynamics in the outer years.

Scott Flanders

Yeah., And let me just add to that, if it answers two questions with one sentence, is we do not expect our gross margin for enrollment to be decremented by any of the market competition. What would cause us to burn more cash is just volumes well in excess of our tailwind case.

Operator

Thank you. And our next question comes from Michael Newshel with Evercore ISI. Your line is now open.

Michael Newshel -- Evercore ISI -- Analyst

Thanks. Can you just comment on the level of conservativism in the guidance since you decided not to leave the full-year revenue guidance unchanged, even though the third quarter came in better, plus the Medicare LTV outlook actually improved and stepped up investment capacity? So if you could just frame that decision, and then also just how you're thinking about outside drivers for the enrollment season.

Scott Flanders

Yeah. So we made the decision headed into the Q2 earnings call to guide once up, and made a decision for a management team and board level that we would not be guiding at Q3. So this is a decision that was made irrespective of where we came in Q3, so nothing should be read into that.

Michael Newshel -- Evercore ISI -- Analyst

Great. Yeah, thanks for clarifying that. And then in terms of upside drivers, then, for the fourth quarter. Since the agent count has doubled, is their productivity the key swing factor? Or is demand for the seniors and the quality of leads also key factors too?

Dave Francis

It's both. It's the volume of customers coming into the platform and our ability to do that on a margin-positive basis from a marketing investment perspective. And then it's the investments that we made on the technology side to make those agents that we've had on board for a while and those that we've onboarded just recently to be more productive, so to in essence increase their conversion rates throughout the fourth quarter. We've built what we believe is an aggressive but very achievable plan with areas of potential leverage in it.

So again, to Scott's point, it's imprudent for us this early in the AEP season to be further commenting on guidance beyond that which we have, but other than to say that we are very pleased with the way that the business has been coming in at, as Scott said, record volumes and record conversions so far.

Michael Newshel -- Evercore ISI -- Analyst

Great. Thank you.

Operator

Thank you. And our next question comes from the line of Lisa Springer with Singular Research. Your line is open.

Lisa Springer -- Singular Research -- Analyst

Thank you. I just wanted to ask a quick question about the expansion of telesales capacity. How much of that was in-house agents versus flexible agents? And is there a difference in conversion rates between those two groups?

Dave Francis

Yeah. The in-house growth was about 50%, so call it 400 up to 600 last year. And then the balance of that is going to be external agents. There is a conversion rate difference.

The newer the agent, the less productive they are on the phone, much as a person playing golf or baseball. The more swings of the bat or club they have, the better they get at using the tools and interacting with the customer. So as I mentioned before, we make a lot of investments on the technology side to make that sales process easier both for tenured agents and for new agents. And we are seeing the benefits of that already this AEP.

But there's no question that a more tenured agent is more productive and more efficient than a new agent.

Operator

Thank you. And I'm not showing any further questions at this time. I would now like to turn the conference back to Scott Flanders for any further remarks.

Scott Flanders

Thank you everyone and for the caliber of questions. It definitely feels to me that the analysts' coverage of eHealth has matured dramatically since a year ago when we were headed into AEP. And I just really appreciate the attention being paid to this name. We could not be more optimistic about how our business is positioned.

4x growth in online, we think, is a harbinger of a sea change in our strategy and our competitive differentiation. And so as I said earlier, we remain uniquely confident in the trajectory of our business, and we believe we're going to have a record fourth quarter. We look forward to telling you about it when the calendar turns to 2020. Thank you everyone for being on the call.

Operator

[Operator signoff]

Duration: 61 minutes

Call participants:

Kate Sidorovich

Scott Flanders

Derek Yung

Daniel Aversano

Dave Francis

George Hill

Jasper Bibb

George Sutton

Frank Morgan

David Styblo -- Jefferies -- Analyst

Greg Peters -- Raymond James -- Analyst

Michael Newshel -- Evercore ISI -- Analyst

Lisa Springer -- Singular Research -- Analyst

More EHTH analysis

All earnings call transcripts