Sirius XM has Serious Underlying Customer Retention Problem

In its latest quarterly announcement, satellite radio carrier Sirius XM tried to sound upbeat. But even as CEO Mel Karmazin tried to point to signs of increased cash flow, the bigger issue became subscriber loss. But the customer count downturn is older than one quarter, and what’s getting lost in the shuffle is just how much auto sales and new subscribers have masked the long-term underlying weakness of the company’s difficulty in retaining subscribers.

As Olga Kharif pointed out in BusinessWeek, investors would bear financial losses so long as new subscribers came in:

Worse still, the company’s subscriber losses may not be over. Many analysts now expect Sirius XM to shed subscribers in the second quarter as well as retail sales remain in the doldrums and auto industry woes deepen. Sirius XM partner Chrysler has temporarily shut all of its factories while it navigates bankruptcy.

However, the issue Sirius XM has with subscription losses has been going on for a while. It’s just not been obvious because of the flood of new subscribers.

For some context, look at the following table, with both Sirius and XM individual rates, as well as combined, over time.

Notice that 2009 Q1 is not the beginning of the subscriber losses. You can also look at the following graph, which helps show the trend. Absent a big change, subscriptions have topped out.

According to Sirius XM, subscriber churn was 2.2 percent in the quarter, compared to the 2008 Q12.1 percent for Sirius and 1.8 percent for XM. And that is true, according to gross measures. Even then, it’s no great shakes. Compare this to the cellular carrier industry. AT&T’s Q1 churn rate was 1.6 percent, down from 1.7 percent in the first quarter of 2008.

But looking at churn doesn’t tell you enough. I went to Sirius XM’s 2008 10-K. Subscriptions via auto sales and leases are vital to the company. However, over the last year or so, those subscriptions starting as promotions only convert into paid subscriptions at a 47 to 50 percent rate. As for those auto buyers or leasers taking a pre-paid subscription, only 41 to 43 percent of them will continue after the initial period. In other words, a majority of new subscribers coming in from auto sales quickly disappear, suggesting that Sirius XM is completely dependent on a frighteningly large influx of new users.

At the beginning of 2008, the company had a pro forma (because it was still two separate carriers) subscriber base of about 17.34 million. It added 7.71 million during 2008, but lost 6.06 million. Consider that for a moment without the additions for a moment. Over the year, Sirius XM lost almost 35 percent of the customers it started with. Compare that to the analogous 32.1 percent in 2007, or the 34.9 percent in 2006. This isn’t a new phenomenon and isn’t the result of slowing auto sales. The company simply can’t keep hold of enough of its subscribers. (And I say that as a long-term subscriber to the XM service.) And if you divide 35 percent by 12 months, that should be 2.9 percent a month. Either my arithmetic is off, of there is some number of subscriber losses that don’t show up in the churn number — perhaps the loss of promotional customers, who never did pay.

Now consider the number of added subscriptions. In 2006, it was 7.63 million. In 2007, 8.08 million. In 2008, 7.71 million. That’s the wrong direction, and becomes a bigger problem as the subscriber base gets larger, because the company needs to add proportionately more subscribers to maintain the growth rates that investors want.

In addition, ARPU — average revenue per user — has dropped from $10.82 in 2006 to $10.61 in 2007 to $10.51in 2008. Now, customer service and billing expenses have been trending downward from $1.31 in 2006 to $1.18 in 2007 and $1.11 in 2008, and so that has more than made up for the declining ARPU. But eventually the savings on service and billing will bottom out, and I’d argue that the declining decrease shows a bottoming out of the possible savings. That means the dropping ARPU will become more dominant and negatively affect revenue.

In short, a problem with subscriber losses seems unlikely to be a concern for just the next quarter. Unless I’m wrong — which is always possible — Sirius XM is precariously perched on the need for customer acquisition, and it is quickly running out of new sources for people who haven’t used the service before.