How much television do we really watch?
I’ve been suspect of the stats coming for years out of Nielsen - especially when it comes to time we spend watching television.
A few months ago, the LA Times reported that Americans were watching nearly 8 hours per day watching TV according to Nielsen. And today, the NYT says that TV viewing is rising still (according to Nielsen).
I’d love to get my hands on that data. We know that Nielsen uses an antiquated paneling system to determine ratings and viewership. That same data is the book by which advertisers and agencies buy time on television. And it drives the entire television economic ecosystem. We need something better. How about a last.fm style/scrobbler for tv?
My suspicion continues big time when it comes to internet use. In the same NYT piece, Neilsen suggests internet use is 27 hours/month while TV fetches 142 hours/month. Internet use is less than an hour a day?
Does that sound right to you?
It does sound right, but maybe not for obvious reasons. The more you know about statistics, the more you can find an angle for a self-supporting thesis. In other words, consider the source of the data and their stake in the outcome ($$). Nielsen is far behind other players in the internet analytics space where competition - including free offerings - creates razor thin margins. It is in Nielsen’s best interest to protect where they have monopoly power (television), and diminish where they never will (internet).
Compare the lengths by which Google protects the authenticity of PageRank to the lack of effort Nielsen puts forth in terms of objectivity. Imagine how silly Nielsen Sweeps Week would be on the internet - every quarter, for one week, every internet site would jam pack their site with their best, most funny, most viral, most salacious content in an effort to appease a statistically insignificant number of paid-to-contribute users. It is absurd that a 24/7 activity isn’t measured on a 24/7 basis with more accuracy. The comparisons get even more interesting when you look at Google’s long tail coverage vs. the scant amount of Nielsen coverage. When you consider how tiny the universe of television is (in terms of concurrent offerings) to the vast internet, and you’ll be shocked at how many shows on TV come up with zero viewers. Really? It is unbelievable because it is not to be believed.
Lastly, funny you should mention a Scrobbler for TV. There already is one in about 60% of homes in the United States. It’s called your digital cable set-top box. If Google ran your cable company, it would be one of their classic “exchange your privacy for convenience” plays - let Google track your viewing activity (live, DVR, what commercials you muted, what shows you turned up, what sites do you search over cable modem while watching particular shows, etc.), and gain from the intrusion by lowering your monthly bill commensurate with the research fee revenue. You’d also gain from some kick-ass realtime aggregation/discovery tools and knowing that you are supporting your favorite shows by watching them - democratization of viewership. I submit, if this were offered in today’s market, Nielsen would be out of business in a matter of months, and we’d see better television programming year round across more channels.
Closing thought: Mechanical Turk + Twitter + set-top box API hack. You see where I’m going with this …
