Last week Steve Jobs presented a new version of the Apple TV device and announced that it would stream only rented content and that TV episodes would be available at a lower price point of $0.99. With these changes Apple addressed weak points in its video strategy, which had caused lackluster sales. However, while consumers may be pleased it seems that the lower price ponit is a point of contention with the television networks. So far only ABC and FOX have signed on to Apple TV, and many speculate that ABC did so since since Steve Jobs is on the board of directors for Disney while in FOX got the OK since Rupert Murdoch is interested in working with Apple on iPad projects that support Newscorp’s publishing business.

Ignoring political alliances, I wondered just what the floor price should be for an Apple TV rental. The blog TVByTheNumbers had a provocative post, hyposthesizing that on average broadcast networks make $0.80 in advertising per viewer during a one-hour broadcast show.

CPM of $25 = $0.03 per view x 32 spots = $0.80 per viewer per episode

I considered this $0.80 should be the benchmark by which a substitution view should be judged. I then set out to estimate the revnues other platforms created on a per viewer basis and compare. The revenue models accounted for the following factors:

• Revenue stream (advertising or user payment)
• Advertisements per episode stream (applicable only to Broadcast TV, Hulu and Hulus Plus)
• Viewers per episode stream (dependent upon the delivery screen)
• Viewings (greater than 1 if the episode has been purchased)

I decided to ignore advertising commissions and distribution fees, which could be a major factor for networks to estimate the benefit, or lack thereof, in distributing their shows via iTunes and Apple TV. At the end of this post are all the calculations for each of the platforms. The following graph summarizes the results:

Even with the higher CPM prices (cited in the Appendix) it seems that Hulu offers the lowest revenue per user. However, Hulu Plus’ dual revenue stream (similar to cable TV), which accounts for a $9.99 monthly fee as well as advertising, jumps the gap and manages to post the highest revenue per user, clearly the winning model. The per-episode purchase models of Amazon and iTunes also pose a decent revenue of $0.66 per viewer while Apple TV’s rental model comes in at a lower $0.50 per customer, which might be too low for certain networks.

As I mentioned before this price doesn’t reflect any distribution fees that Apple might take so the comparison for iTunes and Apple TV might not be accurate. I might also be ignoring other factors. For example, episode purchases through Amazon might have more viewers per episode than Apple iTunes since the Amazon service has more streaming options to televisions.

Another thing to consider is that although broadcast television may take in around an average of $0.80 per viewer during every hour of primetime, these viewers are also likely to continue watching the same channel, contributing to further revenue on the channel during the next show. They are also going to be exposed to TV spots promoting the network’s other content. These are benefits that would might be incurred towards a limited extent with Hulu (which does promote a network and its other content), but would be completely absent with Amazon and Apple. Therefore, there is a much larger set of intangible benefits to a viewer watching a show on a network, that can’t be simply accounted for in the subscriber, rental or advertising revenue of alternate platforms.

I want to also highlight that DVDs are sold after a television show ends, so this aggregate revenue stream doesn’t directly compete with a broadcast audience like Hulu, Amazon or Apple’s services, so it has a different set objectives of to fulfill; not so much to compensate for an audience which is not watching the show on television as helping pay for the production costs- typically a broadcast production only covers 50% to 75% of its costs through broadcast distribution; the rest has to be recouped through syndication and other sources like DVD sales.

Even though I do think these models are inherently limited in their ability to compare the platforms, I still believe the comparison illustrates well some of the factors that are involved in the pricing and negotiation of these deals, and indicate why the new Apple TV rental model is not currently accepted by CBS, NBC and other networks.

Appendix

For Hulu, I calculated the advertising based on a CPM of $63 since according to Wired their rates run about two to three times that of broadcast television. This sounds a bit high to me but it’s possible since the Hulu has better targeting, higher valued demos, and lower commercial skipping/avoidance than broadcast TV.

Hulu: CPM of $63 x 6 spots = $0.38 per viewer per episode

In the case of Hulu Plus I added the $9.99 monthly fee, divided by 18.8 episodes per month (this was estimated by taking the average monthly minutes per Hulu viewer by 47, 44 minutes of content plus six 15-second spots).

Hulu Plus: $9.99 monthly fee / 18.8 episodes = $0.53 in fees per episode + $0.38 in ad revenues = $0.91 per viewer per episode

Since Amazon, iTunes and Apple TV are strictly purchased or rented models, advertising was not part of equation. For Apple TV I assumed that an average of two viewers would watch each rented episode since it streamed to a television. In the case of Amazon and iTunes I assumed an average of only 1.5 viewers since their models could stream to computers that inhibited group viewing. I also assumed that content purchased via Amazon and iTuens would be viewed more than once.

Amazon: $1.99 purchase fee / 1.5 viewers / 2 viewings = $0.66 per viewer per episode
iTunes: $1.99 purchase fee / 1.5 viewers / 2 viewings = $0.66 per viewer per episode
Apple TV: $1.99 purchase fee / 2 viewers / 1 viewing = $0.50 per viewer per episode

The model for DVDs was similar to iTunes and Amazon, only substituting a $2.50 price per episode based on the retail price of $60 for a DVD set with 24 episodes.

DVD: $2.50 purchase fee / 2.0 viewers / 2 viewings = $0.63 per viewer per episode

Last night CBS attracted an average of 106.5 million viewers during its transmission of Super Bowl XLIV, which effectively made it the most watched broadcast program of all time. While the Super Bowl has always been a major broadcast event in the United States its average audience has been growing steadily at around 2.3% since 2000. Last night’s broadcast was the final push, that finally broke the 100 million viewer mark.

This super-sized audience has always commands high ad prices since it offers marketers unique access to a concentrated mass audience. CBS reported that prices floated between $2.5 and $3.0 million, although certain slots went higher. If an average of $30 million is assumed, the effective CPM last night would have been $28.17.

Since CPMs for global viewers during broadcast primetime float around $15, the Super Bowl CPM would be roughly double. This would also be roughly parallel to relationship between broadcast primetime and the Super Bowl within the Household CPM, as reported by the TV Bureau of Advertising.

This is the 100% price premium is the value given by the market towards being able to reach over a third of the U.S. population at exactly the same time. This unique opportunity affords a brand message to effectively become common culture within the span of 30 seconds. Now, whether or not an advertisers use this time wisely is another matter. You can be the judge; see all the spots from last night here at Ad Age.

Last week Ad Age posted the spot prices for the new Fall’s season, based on a “survey of media buying firms.” Combining this information with ratings data for the top 20 programs I compiled approximate CPMs (Cost per Thousand) for the overall audience and the 18-49 demographic.

Most advertisers are not focused on overall audience, since they have far more precise targets for their products or services. In many cases, they also target younger viewers since it’s generally accepted that they’re psychologically they’re less set in their purchasing decisions and more likely to be swayed by marketing messages. This focus on younger viewers has been criticized as “simplistic” by CBS CEO Les Moonves, motivated perhaps to CBS’s dominance of the general audience market. Out of the 20 shows in the following list, 12 are on CBS.

Moreover, popular CBS programs such as NCIS, Criminal Minds and 60 Minutes come up with the lowest CPMs. On average CBS shows have a CPM of $9.76, about 10% less than the group average. Given this access to mass audiences and a efficient broad buy, some budget conscious advertisers opt for CBS:

Donna Spurrier runs Spurrier Media Group, which places ads for companies. One client, Identity Guard, which helps consumers ward off ID thieves, had a small ad budget and chose CBS because it has large numbers of loyal viewers. “When you have a limited budget, you have to fish where the fish are,” she says.

Within the 18-49 demographic CBS’ share of the market is not as strong. The network still lands the most spots, with eight shows in the top 20, while each of the other three networks have four shows each. This information should be taken with a grain of salt however since cue to data limitations for the 18-49 demographic I had to base my analysis on only two weeks of ratings during the month of October. In any case, CBS still seems to be the bargain in this demo, offering a $28.68 CPM, 9% below the average CPM of $31.49.

However, this CPM comparison is by no means the complete story. Advertisers could be interested in much smaller demographic targets which are not being compared above. If they’re targeting younger audiences a show like Family Guy could be more pertinent and more cost efficient. If an advertiser wants sophisticated audiences an investment in The Office or Grey’s Anatomy may make more sense. The latter scenario also ties into the limits of demographic targeting. As I mentioned before, many advertisers are interested in precise targeting, which may go beyond  the limits of age and sex demographics. As one advertiser explained in AdWeek, women 18-34 can mean one thing in New York City and something else entirely in Minneapolis. Therefore, CPM comparisons would be much more refined than what you normally see on a ratings measurement.

Another issue that a standard CPM comparison doesn’t account for is DVR resistance which is why NBC Sunday Night Football ranks so high in these two lists. Sports programs normally assure live viewing thus resulting in a price premium compared to other programs in the same time period. The apt attention of sports viewers is related to another issue affecting TV ad pricing called engagement, which is defined as “how closely viewers are paying attention to the programs.” At a certain level, engagement has already been priced into television slots. Higher attention in primetime is one reason why CPMs are higher there than in daytime. The industry is simply measuring this engagement more precisely now, which may lend it greater weight in commercial pricing standards in the future. But measurement, as they say, is another story.

A couple of articles last week about network television caught my eye. The first, in The Economist looked at how the ‘big four’ broadcast networks have lost ratings during the past ten years. The most dramatic decline has been for NBC; their ratings are been reduced by half. CBS and ABC have fared almost as bad, while Fox has lost about a third. The second article I noted was in Fortune about ‘TV’s Biggest Moneymakers,’ detailing the top 10 revenue generators on broadcast network television. The results are as follows.

What I noticed was how audience size did not completely correlate to revenues. I figured that this was due to two factors: 1) advertisers were not interested in the general audience, and targeting segments like the 18-49 demographic (this explains the revenue powerhouse American Idol) and 2) certain shows with relatively lower audiences, like Private Practice and Heroes, were sold on expectations of higher ratings before their current season began.

Given The Economist article, I was also interested in looking at how network TV ad costs have changed. I found the historical prices for network TV primetime ad spots at the Television Bureau of Advertising’s website, and I also found a chart of Advertising Age’s reports of 30-second spot costs for network TV’s top 10 shows, between 2000 and 2008, at this useful website. Consolidating the two gives us the following picture.

Just as The Economist highlights, ratings have gone down as a whole, and so have primetime ad costs, particularly during the past four years. However, the decrease in ad costs hasn’t been enough to compensate for the decreasing audience, resulting in higher overall CPM’s. It would be interesting to see how the networks will handle the situation with their clients during the  upcoming upfront.

During my last posting I took a look at the audience levels for the Academy Awards since 1974. Since I had the data handy I decided to compare the viewing trends to the Superbowl.

It’s fascinating to see how the audience levels have diverged. Although slightly more popular in 1974, the Superbowl now attracts more than double the audience of the Oscars. Perhaps this is a reflection of the diminishing value of the Oscars, now that there are more opportunities now to see celebrities and movie stars, be it through other awards telecasts, multiple magazines, specialty cable channels and hundreds of web sites.

I was also interested in comparing the relative ad pricing between the two events, especially since it was recently reported in Mediapost that there has been some discounting and special packages offered for advertising in this year’s Oscars. Comparing the ad prices between the two events since 1983, the CPM per household was relatively similar until the last three years.

Last year a 30-second national spot in the Oscars cost $1.8 million and had an effective CPM of approximately $86. This year the prices for a national spot have dropped to $1.4 million. If the broadcast maintains the same ratings, it will have an effective CPM of $66, 23% lower than last year but still higher than the mid 50′s CPM for the Superbowl. Perhaps certain clients will opt to pay more for certain female or upper-income segments that more commonly tune into the Oscars rather than the Superbowl.

UPDATE: The Oscars averaged 36.3 million viewers, well above my predicition and much higher than the audience of 32 million last year. More on this later.

Super Bowl XLIII is less than five days away and I am looking forward more to the commercials than the game itself . A Super Bowl commercial has be better than the game since it has only 30 seconds to grab a viewer’s attention and be as memorable as any play or touchdown that occurs between breaks. This year advertisers will be paying around $3 million for the privilege to air their messages. How did the price get this high and what is it really worth?

The Super Bowl is considered by advertisers be ‘super’ since it manages to draw the mass audience of yesteryear, and it’s still growing. Last year’s Super Bowl XLII between the New York Giants and the New England Patriots yielded a record audience of over 97 million television viewers.

While the Super Bowl audience size has grown, advertising prices have risen at a higher rate. Adjusted for inflation the cost of a 30 second spot has risen 795% between 1968 to 2008. The biggest single annual jump has been 42.5% (in 1985) but overall the greatest growth was seen during the late 90′s, probably an aftermath of the networks’ competitive bidding for the NFL broadcast rights that were then up for renewal. This began in 1993 when fledgling Fox snatched the broadcast rights from CBS with a $1.58 billion bid for four years. CBS then came back in 1998 offering $4 billion for the AFC games, while Fox paid $4.4 for the NFC, which then effectively shut NBC out of the NFL and ended that network’s 59 year broadcast history with the sport. Not until 2006 did NBC get back in the game, by gaining the rights to the Sunday night game, which has subsequently improved its Sunday night ratings. This bidding war demonstrates the power of not just the Superbowl, but of the entire NFL schedule.

But getting back to the Super Bowl, the reason that advertising prices are so high, and the reason that network vie to broadcast the event, is that it’s a unique platform to reach a concentrated audience simultaneously. It’s been shown that Superbowl viewers pay more attention to commercials aired during the event. According to this TNS study, viewers pay attention during the Superbowl at an index level of 100, while the average primetime show has a level of 90. Both advertisers and the networks understand these marketing benefits. Fox invested 44 minutes during last year’s Superbowl for network advertising, a new record. The event also affords a unique opportunity for advertisers to achieve a big splash in the marketplace. It is no wonder that between 20 and 25% of the Super Bowl’s advertisers each year are first-time clients, who probably want to break out of the clutter.

Advertisers will pay premiums for such benefits, which can clearly be seen by the higher CPM (Cost per Mille) for the Superbowl. CPM or Cost per Thousand is an advertising measure often used to measure the relative performance between two media vehicles. The following graph measure the cost of reaching 1,000 homes, either by advertising in the Superbowl or on network primetime television.

The network primetime CPM figures comes from the TV Bureau of Advertising website, and take into account all broadcast networks including the WB, UPN and Univision. Therefore, it might be a little low and not be as different as the Super Bowl CPM, which runs about twice as high. Given the benefits mentioned before, this CPM could therefore be comprehended. While it is difficult to judge whether the prices are worth it, the customers do seem the appreciate the opportunity and that says a lot.