I pulled out the latest EW to cite some concrete numbers, and notice that they list the top 20, not 10. Last week, the top rated tv shows were:
1 Desperate Housewives (22.3 million)
2 Dancing with the Stars (22.0)
3 CSI (20.8)
4 Dancing with the Stars (20.7)
4 Grey’s Anatomy (20.7)
6 Sunday Night Football (19.4)
7 Dancing with the Stars Recap (18.2)
8 Criminal Minds (17.6)
9 Lost (17.1)
10 CSI Miami (16.8)
10 CSI NY (16.8)
12 60 Minutes (16.2)
13 House (16.1)
14 CMA Awards (16.0)
15 NCIS (15.4)
16 Without a Trace (15.2)
17 Extreme Makeover: Home Edition (15.1)
18 Cold Case (14.8)
19 Deal or No Deal (14.7)
20 ER (14.6)
Eliminating all reality shows, sports events, news magazines, and awards shows, we’re left with only 12 scripted shows that generated at least 14.6 million viewers. And 25% of those were part of the CSI franchise. It’s damn hard to get 15 million people to watch something; getting 10 million to download (or order via On-Demand services for a fee) would, IMO, be near impossible.
Your original hypothetical was for a struggling show that couldn’t make it on the networks but had enough quality to be saved by a different model. Friday Night Lights might fit that definition, being a critical darling, but it only brought in 5.9 million viewers last week. (EW noted it in a blurb beneath the top 20.)
While you are absolutely right that the backend model would drive up revenue since the money would still be coming in 6 months later, can you fund a show that way? Does the cast work for nothing but points, and the remaining production costs get paid on credit? Even worse is the advertising. Are you going to spend money on advertising for those six months, or just bank on word of mouth? The latter clearly doesn’t work, as Snakes On A Plane bombed horribly despite unprecedented word of mouth.
Whatever delivery method you use for a subscription model, the consumer will have to do something in order to be able to use it. Do they need digital cable? A computer with broadband internet access? An Xbox Live subscription? Sirius Satellite hardware plus subscription? That requirement put a hurting on Stern’s audience, and he was the top dog in radio for decades.
Also consider that tv shows are very sensitive to lead-ins. If you change nothing about a show except what was on right before it on the same channel, the ratings will fluctuate. ER lost a million viewers when its lead-in was changed from Deal or No Deal to the struggling sitcoms NBC has been pushing. Asking a viewer to go to extra effort to watch a show, and then charging them for their trouble, is going to have an effect several orders of magnitude larger than simply changing the lead-in.
While I love the idea of a la carte cable, I don’t think a pay-per-view model for television is viable. People are too fickle about scripted television to make it worth the risk. Hell, the Sopranos never cracked 10 million viewers, even when you added in all the subsequent re-airings throughout the week plus all the On-Demand orders, and it was arguably the most popular subscription-based tv show in history.
I think commercials are here to stay. Without them, the producers would have to treat every single episode similar to a new movie release. i can’t imagine anyone – producer and consumer alike – really wants that kind of headache.