I like to look at numbers. In the case of Sirius (SIRI), the key figures for me are those that track the Sirius customer activity. They are key figures for Sirius as well. Sirius features them prominently in their quarterly financial press releases.
These numbers are the ones that show how many try out the service, experience the product and continuously decide if they like the content enough to begin to pay for it, to continue to pay for it or to cancel it. It is important to examine the growth of the number of customers that choose to pay for the service themselves because they generate most of the revenue and are the key to driving free cash flow and profits. Sirius calls these customers "self-pay subscribers."
It is no secret that Sirius acquires most of its subscribers by partnering with auto manufacturers and having potential self-pay subscribers experience the unique content through a free trial that lasts from 3-12 months. Sirius refers to these trials as "paid promotional subscribers." At the end of the trial period, Sirius expects to convert a portion of these trials to self-pay subscribers. Sirius reports this figure as a percentage in every quarterly report. In the first quarter of 2011, the number declined to 44.7% from 45.1% in the fourth quarter of 2010 and was down from 45.2% in the first quarter of 2010. The 55% that do not convert to self-pay subscribers are a large portion of the line item listed as "deactivated subscribers."
The other significant figure that affects the growth of the self-pay subscribers is the "self-pay monthly churn." This is the average percentage of the self-pay subscribers that cancel each month. The figure was 2% for the first quarter of 2011, up from 1.9% in the fourth quarter of 2010 and was equal to the 2% in the first quarter of 2010. Note that although 2% seems like a small figure, it represents the loss of approximately 1 million self-pay subscribers in the first quarter.
We know that the total number of deactivated subscribers in the first quarter of 2011 was 1,679,303. This would indicate that the number of deactivations that resulted from those choosing not to convert from paid promotional trials was approximately 680,000. It appears to be reasonable to assume that these deactivations came from the 3,504,165 paid promotional subscribers outstanding at year end 2010, and represents approximately 19% of those paid promotional subscribers.
We also know that there were 2,052,367 gross subscriber additions in the first quarter of 2011, up more than 16% from the 1,720,848 in the first quarter of 2010, although down more than 1% from the 2,075,418 in the fourth quarter of 2010. Although a decline from Q4 to Q1 also occurred at the end of 2009, and although the decline of 1% was far less than the 8.6% decline from Q4 2009 to Q1 2010, I was still surprised and disappointed that the figure declined since I was hoping that the combination of strong vehicle sales, the availability of Howard Stern on smartphones and the push for CPO and other re-activations would spur consecutive quarterly growth in gross subscriber additions.
Sirius subscriber net additions grew by 373,064 in the first quarter, up more than 85% or 201,623 from the first quarter of 2010 and up almost 11% or 44,575 from the fourth quarter of 2010. It is a solid performance and looks like the Sirius management forecasted growth of 1.4 million subscriber net additions will be an easily achieved goal. A careful look shows that self-pay subscribers only grew by 120,844. The rest came from an increase in the number of paid promotional subscribers, or an increase in the number of free trials, by 252,220 to a total 3,756,385. About 45%, or nearly 1.7 million, of these are expected to convert to self-pay subscribers over the next year.
Are these figures important? What do they indicate? Should anyone really care? It most likely depends on how one views the potential growth of satellite radio as a subscription model, the growth of Sirius revenue, profits, cash flow and ultimately the performance of the share price.
Several observations can be made:
1. During the remainder of 2011, Sirius can expect to lose an additional 3 million self-pay subscribers through monthly churn. 2% x 16.8 million (number of self-pay subscribers outstanding at end of Q1 and assumes no growth) x 9 months = 3.024 million. For those that would argue that the churn will decline, the counter would be that the average number of self-pay subscribers would be increasing throughout the remainder of the year. So, if the churn declines by 0.1%, the change in lost subscribers would be 0.1% x 16.8 million x 9 months = .151 million less. However, if the average number of self-pay subscribers increases by .6 million over the next nine months, then the losses due to churn would be an additional .6 million x 2.0% x 9 months = .108 million. One can play around with the numbers, but the difference (.151-.108 =.043) would be about 43,200 for the full year, and should not be particularly significant.
Here are a few more thoughts on churn. First, as Sirius self-pay subscribers trade in their cars and buy new ones with a free trial, there is the potential for an INCREASE in churn. Why wouldn't the car buyer cancel the current subscription and take advantage of the new free trial? Second, churn has been remarkably sticky the past several years hanging around 2% +/- 0.2%. I might have expected that the merger would have resulted in less churn as Sirius and XM stopped going after each other's subscribers. Third, despite the unique content, crystal clear reception, and the relatively low cost of the service (“at a cost of less than $0.50 a day” Mel Karmazin quote during the Q4 2010 conference call), a surprisingly large number of subscribers that have had the service for an extended period of time still choose to terminate the service.
I have read or heard raves from family members, friends and others about the service, indicating that there are many out there that are extremely loyal, have multiple subscriptions and can't live without the service. And even though 2% churn appears to be a small number, it will result in about 4 million self-pay subscribers canceling their subscriptions this year. That's mostly from the 16,686,799 that began the year as self-pay subscribers. How can these views be reconciled? First, perhaps many of these advocates are lifetime subscribers and never appear in the numbers. Second, there is most likely a strong core of subscribers with extremely low turnover, only canceling when they buy a new car, and the remaining self-pay subscribers have a very high turnover.
2. Satellite radio has been compared to the subscription model for cable (I will use the term cable as a generic reference to include co-ax, fiber and satellite) TV, and since cable TV has become ubiquitous, satellite radio will follow the same path. I just don't see it. Cable subscribers fell in love with the product and kept paying. This is not what is happening with satellite radio. There are going to be those out there that scream that cable TV has a higher churn rate than Sirius. Maybe it's true and maybe it isn't, but churn is not comparable from one company to the next. Some companies count trial cancellations in the churn figure (Sirius doesn't) and some report it as a quarterly figure or at an annual rate (Sirius uses a monthly rate).
Cable TV churn rates should be higher than satellite radio. Sirius and XM merged and eliminated direct satellite radio competition. The landscape is different for Cable TV. The junk mail cluttering your mailbox or ads one sees on TV for FiOS, DirecTV (DTV), Dish (DISH), Comcast (CMCSA) and others are testament to the marketing battles taking place and evidence of these service providers going after each other’s customers. Furthermore, when someone moves from one home to another, they will cancel their cable and get a new subscription from the local service provider. (This is unnecessary for a Sirius subscriber.)
Think about college students or young professionals changing apartments. Anecdotally, my son is going into his fourth apartment in four years of college/grad school. Different names go on the bills and they sign up for the double play - high speed Internet and cable TV at the "low" introductory rate. Then there are those cable subscribers with two homes that might be turning the service on and off as they spend summers in the north and winters in the south. It drives up the churn rate for the companies involved. In the end, though, cable subscribers are the dominant audience for TV reception, dwarfing broadcast, or free, TV.
As I pointed out earlier, I expect that turnover of cars will eventually drive churn higher, but I believe that this is not yet significant. I have absolutely no proof that this is the case, just a hunch. It would mean that if the current OEM arrangements continue, at some point Sirius will be hurt from both a revenue and a cost perspective. They will be subsidizing the cost of the radio in the new car, receiving less revenue from the paid promo sub than they were receiving from the trade-in vehicle and sharing more future revenue from the new vehicle with the OEM.
3. The 3 million self-pay subscribers that will cancel (based on 2% churn) during the remainder of 2011 need to be replaced. Most will need to come from the 3.8 million paid promotional subscribers outstanding at the end of Q1, plus some of the shorter paid promotional trials (those less than 9 months) that get added during the remainder of the year. Others will likely come from reactivations, CPO, smartphones, retail, Internet subscriptions, etc.
So, how does this work? Well, if we start with the OEMs and look at a figure of 13.5 million new vehicle sales, and a 62% penetration rate (derived from statement by Mel Karmazin in the Q4 conference call: “And our penetration level has been picking up a little bit actually, so we're in the low 60% and we don't see that changing.”, we get about 2.1 million gross subscriber additions per quarter. Even 14.0 million new vehicle sales and 65% penetration would only bring the average figure up close to 2.3 million gross subscriber additions per quarter. We also know that about 46% of the paid promotional trials converted over the past several years and that Sirius management does not see that figure changing significantly.
Okay. Let's say Sirius can get 46% of the 3.8 million paid promotional subscribers outstanding at the end of Q1 to convert by year end. That means they get 46% x 3.8 million or 1.75 million self-pay subscribers over the next nine months. They still need to add another 1.25 million self-pay subscribers just to make up for the losses from churn over the next nine months, and then they still need to increase the number of self-pay subscribers. Since Sirius has been able to accomplish this for the past few years, it would indicate that many of the trials are for shorter durations and will be coming from the 6+ million OEM paid promotional trials to be activated over the rest of the year, as well as the new sales from their other channels.
4. Lots of people just won't pay for the Sirius product. These aren't people that have never heard of the product or experienced its exclusive content, commercial free music, or crystal clear, uninterrupted service. These are the 11,000 self-pay subscribers that cancel every day. These are the 55% (about 4.5 million) of new car buyers in 2011 that get a free trial and will choose not to become subscribers and pay for the service.
For some, Sirius is a necessity. For some, Sirius is an affordable luxury. For most, Sirius is a discretionary item. It is a discretionary item that a lot of people choose to do without. The numbers speak for themselves. 6+ million decided to do without this affordable luxury each of the last three years, and there is a good chance that figure will increase to as many as 7 million in 2011. That's more than 25 million subscribers that chose not to pay for the service.
5. What should the cautious investor be looking at with respect to subscribers? First, monitor the conversion rate. It has continued to slide. Is it the economy? Sirius is working with the OEMs to place radios only in those cars that have the highest likelihood of converting to self-pay subscriptions. The conversion percentage should be rising and it isn't. The 5 most recent quarters are:
Quarterly Conversions to self-pay
Also, continue to look at the churn.
Finally, it should be noted that Sirius subscribers are not necessarily unique individuals, but rather represent individual radio subscriptions. According to Sirius Investor Relations, Sirius counts each subscription as a subscriber. If an individual has two cars, each with a satellite radio, and listens over the Internet at home, even if all three are billed as a family plan on the same invoice, it is three subscribers.
Additional disclosure: I have written covered calls at various price points and various expiration dates against most of my SIRI position. I also have positions in VZ and CMCSA which are referenced in this article.