Analysts are expecting Sirius 2006 revenues of approximately $615 million on about 6 million subs (up from 3 million prior to Howard's joining). Similar to the simple-headed but somewhat effective method I used when justifying Google's purchase of YouTube on a valuation basis, I will seek to bring some back-of-the-envelope intuition to the potential for Sirius' ad revenues.
2 channels x 24 hrs/day = 48 hrs of programming/day
x 3 ad minutes/programming hour = 144 ad minutes/day
x 200 broadcast days/year = 28,800 ad minutes/year
x $5000-$10,000/ad minute = $144-$288 million ad revenue/year
Wow. Is this possible? Haircut these numbers by 50% and it still looks like Howard has the potential to cover his basic $100 million nut based upon ad revenues alone. Add in the value attributable to those incremental subscribers (3 million x $120/year = $360 million) and Mr. Clayton's pie-in-the-sky deal doesn't look so crazy after all. Further, it appears that Howard's Sirius' ad revenue numbers could eventually eclipse those from his Infinity days.
Further, there are some very interesting discussions concerning the technology issues embedded in the merger, i.e., whether SIRI and XMSR users have the ability to get both signals through their existing receivers, the ramifications of large numbers of terrestrial repeaters within key markets across the U.S., etc. These issues, while important, have certainly not been in the forefront (or even holding up the rear) of discussion in the wake of the merger announcement.
After collecting some additional data and attempting to connect the threads running across several important themes, I'd like to put forth an even bolder vision of the merged entity:
A subscription-based provider of premium digital content, with the ability to deliver targeted, relevant advertising through the airwaves as Google does over the Internet.
I know, this is a pretty bold vision of the future for SIRI/XMSR. Let me first validate the thinking around advertising as a key value driver of the merger, and then proceed to the somewhat more out-of-the-box thinking relating to targeted advertising through the airwaves. I've got to think that a big-picture, swing-for-the-fences, deal-making and advertising-savvy guy like Mel Karmazin has to be thinking about this stuff; he's just keeping it to himself in order to shock the world. But not me.
Advertising and The Merger
Howard and Mel had a little chat about the merger at 7:30am this morning. Two big guys, two big personalities. Good stuff. Mel raised a lot of interesting points during the discussion, which you should definitely check out, but the key point related to advertising and governance is here (from MarksFrigging.com):
Mel said that each of the companies have advertising and it will be a good thing when they can go to advertisers and tell them that they have over 13 million subscribers to advertise to. Howard asked Mel where they were holding these meetings and stuff. Mel said he had them over to his apartment and that's where they would go over all of this stuff. He said that things didn't work out for a very long time because there were arguments over who was going to get 55 percent of the company and who was going to get 45 percent. Mel said that XM was arguing that they had more subscribers and more deals with car companies. Mel said SIRIUS argued that they had better content and were growing faster. Mel said that at the end of the day it was a 50/50 deal. Mel will be the CEO of the company and Mel said that both boards agreed on that whole thing. He said he would have been happy to step aside if that wasn't acceptable to them.
13 million and rising fast - that's a pretty serious user base off which to work. It's also interesting to note that a little visit to the SIRI website turns up seven recent job specs for Account Managers selling radio advertising. Mel knows that while the subscriptions are nice, advertising is the big embedded call option an investor is getting in the merger, an option which he fully intends to monetize for the benefit of SIRI/XMSR shareholders - including himself.
Another point I'll make before getting into the technology discussion is that each satellite receiver has its own unique digital ID. So what? Well, presumably that unique ID is attached to a person or people that have preferences - preferences which can be gleaned by usage (i.e., this ID listens to these shows at those times and in these locations) or by selection (i.e., checking off interests on a secure website which then facilitates targeted advertising based upon user profiling). All I'm saying is that this is an interesting factoid that warrants a little consideration - and expansive thinking.
Advertising and Technology
Terrestrial repeaters. Huh? A little background, from CED Magazine 12/01/2006:
In 1997, XM Radio and Sirius each won satellite radio broadcasting licenses in an FCC auction, each paying around $85 million. The FCC called the service Satellite Digital Audio Radio Service (SDARS). From the start, SDARS was expected to need a network of terrestrial repeaters to fill in shadowed areas.
This was flagged as a "hot button" issue between the satellite radio companies and the terrestrial radio companies (as represented by the National Association of Broadcasters - or "NAB") several years ago. From audio/video revolution 06/30/2004:
Both satellite players use terrestrial repeaters (similar to cell phone towers) in major cities to bolster reception of their signals. A spokesperson for Sirius tells AudioRevolution.com that Sirius has terrestrial repeaters positioned on top of buildings, including the McGraw Hill building in Manhattan and on top of a medical building near the 405 freeway in West Los Angeles. In the case of Sirius, these repeaters reach five to 10 miles in range and help broadcast Sirius’ signal into the hard-to-reach canyons of Los Angeles, as well as into the cement super-structures in New York.
Some repeaters have the potential of more power – a lot more power. Several years ago, reports from Inside Radio, a leading radio industry trade publication, said that some of XM’s terrestrial broadcast capabilities in cities like Boston are as much as 50,000 watts, meaning they could compete with the strongest FM stations and have exponentially more power than a smaller repeater. Thus the concern that has the radio industry up in arms. Under this scenario, the satellite providers are just an FCC ruling away from being competition for the local ad dollar. This would be catastrophic for terrestrial radio, which depends largely on local, not national, revenue to make its profit.
No less than the radio trade organization NAB was an early accuser of satellite radio’s intentions. Powerful chairman Edward O. Fritts came right out and said it – satellite radio wants to get into the local radio business. While XM and Sirius denied it, there is a general feeling among radio broadcasters that they would not be surprised to see their satellite competitors try to win regulatory changes that would allow such competition.
What traditional radio does best in the world of advertising is reach customers locally and regionally. The satellite providers are starting to legally provide traffic for major cities, which possibly points towards their interest in providing more local content for their users. The traffic move could also simply be a way to keep their new and existing subscribers from having to flip-flop back to terrestrial radio to get critical traffic information. This is not dissimilar to TV satellite operators. When satellite television was able to provide local channels, new subscribers signed up by the millions. The satellite music companies would likely to see the same kind of bump. They also could sneak in local ads that could potentially generate hundreds of millions of dollars of revenue.
So you get the point. But what of this move to potentially attack the local advertising markets? Back for just a moment to our friends at CED:
Those poor folks at Sirius and XM Radio. They just get finished with one FCC scandal–overpowered FM modulators that violate FCC Rules–when they have to disclose that their networks of terrestrial repeaters have been operating illegally. As it turns out, those violations are no big deal. But as I began to investigate this latest controversy, it suddenly hit me. They aren't really satellite broadcasters with terrestrial repeaters to fill in the coverage gaps. What they really are is terrestrial broadcasters with satellites to fill in the coverage gaps.
You can hear the terrestrial radio broadcasters right now shouting "Danger, Danger, Danger..." Now if you really want to adopt a "conspiracy theory" mind-set (e.g., the satellite broadcasters knew all along that this was the endgame, regardless of what they've said in the past to the FCC and the NAB), check out this link showing the Sirius control room. See that screen in the upper left? Those are the locations of its terrestrial repeaters. Notice the bunching in those heavily populated, super-valuable advertising markets? They've got local broadcast capabilities even through they're a "satellite radio company." Ha! And the punch line from CED:
From a spectrum policy perspective, there should not have been any need for XM and Sirius to register the locations of their repeaters with the FCC. They both have exclusive nationwide licenses for their frequencies. If they cause co-channel interference, it would be only to themselves. If the FCC gets around to adopting rules for these repeaters, the rules will undoubtedly say that repeaters can be deployed anywhere, so long as they meet the as-yet-undetermined technical rules. But under the FCC's temporary authorization rules, every transmitter must be registered. So this "scandal" is much ado about nothing.
On the other hand, this controversy has revealed that XM Radio and Sirius have cleverly transmuted their satellite licenses into terrestrial broadcasting licenses. Sure, the NAB is unhappy. But there are millions of listeners who love it...and most of them don't even listen to Howard Stern.
The whole point of this line of discussion is to make the following argument: the real upside in the SIRI/XMSR deal is the advertising, not the subscription revenue. I think some of the biggest questions are:
- How quickly can an integrated advertising strategy be developed?
- How quickly can SIRI/XMSR users get access to both signals across a single device?
- How quickly can ad rates reflect the combined subscriber base versus each in isolation?
- How quickly can metadata be captured, collected and monetized based upon listener behavior and preferences, akin to cookies and/or memory of historic activity?
Getting the answers to these questions is, without question, worth billions in equity market value to the combined entity.
There has been a lot of discussion around the ability of the existing satellite receivers to pick up different signals, i.e., of SIRI and XMSR. The SIRIUS Backstage blog had a pretty comprehensive discussion of this very issue on 02/21/2007 in a post titled "Can Sirius and XM Radios Pick Up Both signals today?"
While there is no concrete answer, we have found a couple pieces of evidence that Sirius and XM may be able to keep the current radios of today and still offer all of their content to both sides. First, at Select Satellite Radio, the group that is in charge of making a dual-radio in order to ensure the FCC stipulation that they develop one, states this:
“It is acknowledged that SIRIUS, XM and their manufacturing partners already produce receivers that permit end users to access all Satellite Digital Audio Radio systems in compliance with FCC interoperability obligations.”
It appears that this implies radios today can receive the entire satellite digital audio radio services(SDARS, the technical term for satellite radio) spectrum.
So, right now we have two questions that need to be answered:
1) Do XM’s newer radios still decode PAC?
2) If so, does it mesh with the evolved version of PAC that Sirius uses?
If the answer to both is yes, then it looks like there is a strong chance radios will be able to pick up both services down the road. This of course assumes that the merged company broadcasts solely in the PAC codec.
Ok, sounds pretty good, right? But in the comments to the post, there are some amazingly detailed and educated perspectives on precisely this issue, with the conclusion being that the existing receivers are not sufficient to pull down both signals. Check out comment #18:
I'm an engineer, so I know a bit about how the radios work. Without getting too technical, the best likely case is that an existing Sirius or XM radio might be capable of receiving both sets of signals, but they'd always be handled as two different sets of channels with a "hard" mode switch between them (kind of like "AM" vs "FM"). In other words, you could switch your radio from "Sirius" mode to "XM" mode, but it would take a few seconds (maybe 20-50 seconds or more) before it locked on to the new mode's signal and authenticated itself.
That's the best case. I'm not saying it's definitely possible, just that even under the most ideal circumstances, the best anyone with an existing radio can hope for is to be able to reflash their radio and switch it back and forth between "Sirius" and "XM" modes.
OK, I lied. I'm going to get techincal anyway. Here's the hardcore explanation:
Sirius radios don't tune channels. They tune 3 bitstreams -- two from the satellites above, one from the nearest terrestrial repeater. The signals are combined together to produce a single encrypted bitstream that carries all of the audio channels and meta information about them. This encrypted bitstream gets passed along to the baseband chip.
The baseband chip checks its records to see whether it's been authorized to decrypt the bitstream. If it hasn't, or it's about to need a refresh, it decrypts and examines the first chunk of the encrypted bitstream's datagram for the subscriber ID burned into the chip at the factory. If it sees its number, it notes the authorization. Otherwise, it waits for the next datagram and does nothing further with the current one.
Once authorized, the baseband chip deserializes the bitstream into the various channel streams. The desired stream then gets buffered and fed to the codec, which transforms it into an uncompressed PCM datastream. That datastream passes through a low-end DSP, and finally gets fed to a digital amp, which is basically a high-power DAC whose output needs no further amplification. Alternatively, it outputs it to the FM transmitter for PnP use.
Anyway, the point I'm trying to make is that the chips involved are fast enough to process Sirius' bitstream in one gulp, but NOT fast enough or equipped with sufficiently-large buffers to take in a bitstream twice as large (encompassing Sirius and XM's content) in a single operation. They're pretty fast chips to begin with, and value-engineering will ultimately result in chips that are just barely fast enough to do the task they're designed for, but nothing more.
So maybe a quick fix to this receiver issue isn't in the cards. No matter. The potential value we're talking about far dwarfs the costs associated with either a retrofit or a receiver replacement program. Why? It's all about the data.
Google in the Sky
When the services are integrated, you'll have a broader array of programming available than if you were a subscriber to only one. This may encourage more people to enjoy the benefits of high-quality, crackle-free, proprietary content available from the merged entity. And even before more subscribers join, SIRI/XMSR will have 13 million+ legacy subscribers from which to entice advertisers. This is real power. But wait; remember that stuff we were talking about concerning the ability to deliver targeted ads locally (terrestrial repeaters dotting the densely populated, high-value landscape)? And remember the unique ID that each receiver has? And you know how SIRI and XMSR "push" information to subscribers, whether it is the name of a song, an artist, a sports score, the url of an advertiser (in the case of SIRI), whatever? What if they started to "pull" information concerning station choice, location, time of day receiver is active, length of time receiver is active, maybe user-provided preferences concerning various interests (restaurants, concerts, merchandise, etc.)? What's to stop them from delivering tailored advertisements and content recommendations to you, a la Google and contextual advertising? All the pieces are there. It is simply a matter of putting it all together.
Does this sound crazy? It doesn't to me. Now there is the small matter of the Justice Department and the FCC letting the merger happen, which is kind of the precursor to this whole line of discussion. But assuming the merger does go through, and SIRI/XMSR captures the benefits of the combined cost savings together with a larger and more attractive subscriber base, it will have the resources necessary to access, analyze and monetize the user-generated metadata to turbocharge their advertising machine. And this could represent a new paradigm for how value is extracted from the airwaves. Kind of like how Google showed us how to extract value from the Internet.
Am I dreaming? Maybe. But sometimes dreams come true.
Disclosure: Author has no position in stocks mentioned.