The subject of subscriber numbers can arise in many contexts. Depending on the debate at hand, how subscribers are counted can be good, bad, or indifferent. Depending on the perspective, Sirius and XM differ a bit in when subscribers are counted in the OEM channel. Both count subscribers for which they receive payment. The functional difference is when the payment is received from the larger OEM partners of these companies. For Sirius, payment is received when the vehicle is produced. For XM, payment is received when the vehicle is sold. Thus, the pipeline differs between the two companies.
The XM method is ARPU friendly. The subscriber is an end user and the dollars are translating to the bottom line and ARPU. The Sirius method is not ARPU friendly. Although Sirius has been paid, any cars that are not yet sold ("parking lot subs") have the revenue associated with them in deferred revenue. Deferred revenue is a liability. The payments from the OEM will not become revenue until the car is sold. For Sirius, the "parking lot subscribers" represented about 900,000 (11%) subscribers out of the 8,300,000 subs on the Sirius books. Because of this, the ARPU number is impacted by 900,000 subs that are counting as $0 towards ARPU.
Cash flow is very important to SDARS companies. In many ways, the Sirius OEM deal structure is very cash flow friendly. Sirius pays a subsidy to the OEM, and then as soon as the car is produced, Sirius gets a subscription payment. This payment can then be used to get more OEM units installed. While the OEM channel builds, the self paying subscriber pool will build. Monies can quickly be re-invested into the OEM channel to facilitate the process.
The XM method is not quite as cash flow friendly. They subsidize the install, and then wait for a period of time for the car to sell. With GM, the recouped revenue from the OEM is two months worth of service. The smaller term of the promotion converts subscribers to self paying faster though.
Ryan over at Orbitcast had done an interesting article, Why Arbitron's Satellite Radio Ratings Are Wrong, about Arbitron ratings for satellite radio. In the piece he pointed out several flaws in the Arbitron system. Ryan Saghir also did an Orbitcast radio show on the subject. Piggy-backing on his analysis, I would also consider that the Sirius method of counting subscribers would tend to flaw the Arbitron process as well......At least in as much as people try to compare the ratings with subscriber bases that are almost equal. Arbitron is measuring "LISTENERS" not subscribers. The LISTENER number for Sirius and XM differs by a margin greater than people think. While the difference in subscribers is 700,000, it could be argued that the difference in listeners is more along the lines of at least 1,600,000. Confusing matters further, there are LISTENERS who are not yet counted as subscribers. Toyota, Nissan, Subaru, Hyundai, Kia, Porsche, BMW, Mini, and a few other OEMs all have listeners that are not counted as subscribers until the LISTENER becomes self paying. This is because neither Sirius nor XM receive subscription payments from these OEM's. These OEM's buyers are "PURE TRIAL SUBSCRIBERS", thus there are indeed many LISTENERS who are not yet counted in the subscriber numbers for either company.
If potential advertising clients look at the Arbitron numbers, it will become harder for Sirius to sell ads. The have less listeners. A potential ad buyer will naturally compare the two services when considering advertising, and simply stated, XM has more LISTENERS. It would be of little surprise to me if the ad sales staff of Sirius had a tendency to concentrate on the subscriber number rather than the listener number or the Arbitron number. This gives XM an advantage when an advertiser simply wants to reach as many people as possible.
Sirius tends to have longer OEM terms. Thus, they are able to hold LISTENERS for a longer period of time than does XM. In this aspect, Sirius is helped not only by keeping a subscriber longer, but by keeping a LISTENER longer as well. Sirius method is more churn friendly than XM's. XM gets a trial sub for three months, and then about half of those SUBSCRIBERS/LISTENERS churn out. With Sirius, they are able to hold the SUBSCRIBERS/LISTENERS for six months or even a year.
The lesson here is that each company has slightly different ways of counting things, and that each method will either help or harm the company depending on the discussion at hand. Comparisons between the two can be fraught with speculation, and assumptions. With the first quarter conference call happening soon, we will discuss some of these dynamics on Sirius Buzz Radio this Thursday May 8th at 10:30 PM. Readers can also listen to our previous shows at the same link.
Position - Long Sirius, Long XM