Why Good TV Shows Get Cancelled

On December 16, 2010, SyFy Channel canceled its science-fiction series "Stargate Universe", shortly after the airing of its midseason finale (1). On the surface, the news is not surprising. "Universe" was presented as a "darker and edgier" spinoff of the venerable science fiction franchise, but it didn't click with the cable viewership. Despite an impressive 2.35 million viewers on its series premiere, the number dropped slowly over two seasons; by the season two opening the viewership had eroded to 1.175 million, and just over 1 million watched the midseason finale, according to Television by the Numbers (2). Canceling shows that have lost a chunk of viewership is the usual operation of any television channel, broadcast or cable.

The numbers and metrics of who is watching "Universe", however, gets far more complicated than the simple figures above suggest. The vast majority of viewership numbers are provided by the venerable Nielsen trackers. According to their site, Nielsen has been using devices that track what viewers are watching and when they are watching it since the 1930s.

Nielsen ratings are what everyone uses, and so they are important. Especially important is the Nielsen's report on each show's demographic performance. The coveted 18-34 bracket is what advertisers market to, so show X can have lower overall viewership than show Y, according to Nielsen, but still make far more selling ads because it has a higher share of the key demos (3). But the problem is that it's not feasible for Nielsen to track every television. So they cheat.

The result is that Nielsen has a very small sample of households actually tracked for data measurement-historically less than one percent. Furthermore, the pool isn't a random sample, meaning that statistically speaking it is very suspect to generalize responses from the Nielsen households to the entire national population. The result has been that every television show is fighting numbers that may not be anything like the true number of people watching the show.

But that has been the case for decades. The problem is that while Nielsen's flawed methodology has remained constant over the years, the television landscape has dramatically changed. The first big difference from, say, thirty years ago and today is that cable television rules. Years and years ago, the television-watching nation was a much more uniform demographic, watching the same core set of shows. Now, more people with different tastes watch television, and at the same time there is far more choice now then in the days of the Big Three networks. With hundreds of channels to choose from, there is a need for more sophisticated ways of measuring viewership, but Nielsen has been slow to respond.

The second major change has been the advent of delayed watching. DVRs make recording live television to watch later very easy, and as DVRs have become commonplace, the "live" viewership has eroded. Fairly recently, Nielsen came up with an expanded definition of people watching DVR'd programs "same day", and within a week of broadcast. These expanded brackets change the viewership numbers significantly; in the case of "Universe", it often added an even million to their number.

Not that it mattered, of course. This is the failing of the Nielsen ratings; because DVRs can theoretically skip ads, the inflated numbers do not matter for the purposes of whether or not to cancel a show, a decision based not on the audience per se so much as ad sales and production costs. What viewers don't often realize is that by changing their viewing habits, they aren't being counted in the same way and for Nielsen viewers, that means that they are essentially hurting the shows they enjoy watching.

The problem means that shows good and bad, even ones with positive critical reception and good-sized audiences, will be cancelled because their audience does not tune in the same way. Did you buy a season pass for a show on iTunes? The studios are not making the same kind of money on those, so your consumer "vote" does not count. Stealing a show by downloading a torrent is, of course, another way that these shows lose viewership. The upshot of all of this is entire markets are essentially cannibalizing their own programming. The future of television is not looking good for either the networks or viewers.

The solution is a give-and-take. Nielsen needs to step up its measurements and deliver more accurate and data-rich results. Advertisers need to figure out a way to have a better chance of their ads reaching the eyeballs they purchased space to reach. And the future of profit-sharing for online ventures such as Hulu needs to be approached in a new way. Viewers are used to getting things for free, and with such an ingrained dynamic in place networks and providers need to realize that working with the current system rather than fighting the tide is the best way to a happier consumer and a bright economic future for everyone involved.

References
* (1) Kate Ward (December 16, 2010). "Stargate Universe Cancelled". Entertainment Weekly. Retrieved December 30, 2010.
* (2) See the references linked from "List of Stargate Universe episodes".
* (3) Brian Steinberg (October 18, 2010). "'American Idol' spots still priciest in Prime Time". Advertising Age. Retrieved December 31, 2010.