Retransmission and affiliate fees, mostly set by TV ratings, are keeping revenues per viewer high, despite the fact that recovering ad spending remains well below historic levels. The growth of these fees is especially nudging television toward a digital future of more diverse income and rescuing broadcasters from financial crisis.
The $25 billion in cable affiliate fees paid in 2009 was double national broadcast advertising revenues and outpaced cable advertising, according to an intriguing new report by Barclay's Capital analyst Anthony DiClemente. Ratings-based cable networks collectively received an average annual $1,500 per viewer in affiliate fees, as estimated by the analyst over a three-year period.
Cable networks owned by Walt Disney Co. and News Corp. generate disproportionately higher fees for their popular live sports. Pure-play cable networks, such as Scripps Network Interactive's HGTV and Food Network, dramatically underdeliver fee revenues from their loyal niche audiences, providing plenty of room for growth.
advertisement
advertisement
Overall, the blended affiliate fee for cable networks is 27 cents per network. An average target audience of about 200,000 for the total day yields about $1,500 in average affiliate fees revenues per viewer per year, according to DiClemente's math.
The ad-dependent big four broadcast networks, for which retransmission fees have become a frothy new revenue stream, are generating about $6.30 per subscriber per month, or nearly $3 billion annual revenues from retransmission fees -- representing plenty of upside.
While the analyst clearly demonstrates rising viewer values at a time when television is widely perceived as a digital nonstarter, the economics continue to be based on Nielsen's chronically flawed ratings system. Nielsen's audience measurement system is rooted in estimates based on sampling. It smacks of an archaic process in an era of burgeoning one-click accountability. Still, Nielsen's ratings are the way television conducts its business. Average revenue per viewer represents the most consistent comparative metric as it normalizes for the number of subscribers per cable network, DiClemente says.
Fast-growing retrans and affiliate fees represent the most significant trend in television; they provide a robust way of generating new revenues from existing programming. This also sets up TV content providers for the pay-to-play business model morphing all around them, whether it is more than 1 billion Apple iTunes downloads or the 55% of Netflix subscribers now streaming movies.
Drilling down into the economics of retrans and affiliate fees renders other interesting insights about the slowly transforming television sector:
*These shifting economics make a case against a la carte pricing because of the implicit need for "non-watchers" to subsidize "watchers." Television continues monetizing bulk content for the "masses" even as the Internet and connected devices focus on generating premiums for reaching individual targeted consumers.
*These fees will be critical for smaller, more traditional ad-based media concerns such as CBS, where an incremental $750 million of retran revenue, at a 90% profit margin, yields as much as $675 million in earnings, or more than one-third of CBS' 2010 estimated $2.2 billion EBITDA (earnings before interest, taxes, depreciation and amortization).
*As ad dollars continue to be diffused across the digital media spectrum, program networks will continue to press for new ways to analyze and capitalize on viewer value. The shift was evident in 2009 when cable affiliate fees topped $25 billion, dominating nearly $19 billion in DVD/home video revenues, nearly $18 billion generated from cable advertising, $13 billion in broadcast network advertising and nearly $11 billion in box office receipts.
*As TV program viewing becomes more diffused across mobile Internet devices, aggregate viewership (rather than the broadcast network's appointment prime time) becomes more meaningful and will reset the economic inequities between cable and broadcast TV.
*There are at least three potential sources for increased incremental fees for content owners. Since net video subscribers did not decline during the recession, DiClemente says consumers "could potentially tolerate further price increases" -- if you believe that they pay for entertainment as if it were a basic living necessity. Cable distributors sporting an estimated 55% gross video margins this year can afford to pay future fee increases for must-have-programming.
*Independent, non-sports cable networks, such as SNI's Food Network, Viacom's Comedy Central and Cartoon Network and Discovery Communications channels, provide cable operators "the biggest bang for the buck." They generate fee revenues per viewer that are generally more than 50% below the industry average. But that, too, will eventually change.
I must take issue with such ill-informed comments as "Nielsen's audience measurement system is rooted in estimates based on sampling. It smacks of an archaic process in an era of burgeoning one-click accountability."
I am in no way defending Nielsen (full disclosure: Nielsen Australia employee 1977-1997), but as a statistician I am defending sampling. The day we walk away from sampling as a reliable and robust statistical procedure is the day that the lunatics have finally taken over the asylum.
If I may give you one example where a sample is not only representative, reliable and robust, but preferable - a blood test. I sure as hell don't want the doctor conducting a census and draining me dry.
If I may also give an example comparing sampling with 'one-click accountability'. Here in Australia (with a population of just over 22m) the TV ratings are conducted (by OzTAM) using technology and sampling procedures similar to Nielsen in the US. The TV data show usage of around 14m a day, 18m a week, 21m a month and 22m a year (all data rounded) - entirely believable data.
In contrast the online 'one-click' data - based on cookies (the bedrock of the majority of on-line measurement) - now shows that in Australia 89m people are online every month. Yes, that is FOUR times the entire Australian population - and obviously a totally unbelievable figure. (Interstingly, the sample based approach shows 15m - far more believable given that internet penetration is not 100%).
The era of sampling is alive and well. The era of cookie-based one-click measurement is clearly EXTREMELY ill - and getting sicker every month.
Diane, your article is "dead on" please keep us posted with more info regarding this subject.