Nielsen Might Be the Reason Terrible Shows Stay on the Air

Illustration for article titled Nielsen Might Be the Reason Terrible Shows Stay on the Air

Not to beat a dead horse until it's the deadest, but it'd be great to learn the reason why Dads hasn't been cancelled, besides nepotism and the continued power of the white male in an industry supposedly dedicated to creativity. The show's ratings haven't even been that fantastic, though they were better than some of its freshman counterparts (RIP, We Are Men). As hard as it is to admit, America, people are watching Dads: a whole 3 million people watched last week's episode. You can, however, take comfort in the fact that that number could be worth absolutely nothing.


Much has been written about the fear the internet has struck in the hearts of network executives; as the ability to watch television not on the actual TV set has expanded, their ability to figure out exactly who is watching what has decreased dramatically. The bottom line is that the internet has made everything less neat. But there may be another culprit to point to in this numbers game: Nielsen, the company that measures who is watching what. Exhaustive research done by the Council for Research Excellence, reported on Poynter, indicates that Nielsen's not casting a wide enough population net in gathering their television ratings, and the information they are getting has an increasingly high error rate.

Nielsen ratings are collected in one of two ways: diaries, which requires participants to write down what programs they watch and report them to Nielsen every so often and meters, a box attached to the television that collects that data. Diary collecting is problematic because self-reported data is rife with flaws. Meter reading is problematic because we live in a world with Roku, where people watch television on cell phones. A surprising 30% of Nielsen's data still comes from diary entries, and the large percentage of people that fill them out look how you might think they would: white, middle-aged or older, with higher education levels. They watch basic channels: ABC, FOX, CBS, NBC. They have a landline. They are not actually what America looks like. They're what the GOP thinks America looks like.

The people who are being underrepresented in this data collecting, in both diary and meter readings, are what America does look like. They're black and hispanic, they're younger, they rent their apartments, they have more kids, they use cell phones. They watch television online, channels like BET, MTV or HBO. They watch TV in groups.

This data divide is problematic because it means that Nielsen is essentially not reporting data about a percentage of the population that accounts for a big portion of where the money is, that important 18-49 year old young, hip bracket. Advertisers pay more to advertise during shows that do well in this age range, so if those people are being underrepresented, that means that the shows they do watch, the ones that are likely a better representation of what they look like, don't make as much ad money and have a higher chance of being cancelled.

It also means that the creators of shows that actually include people of color or women or LGBT characters or – gasp! – someone who fits all three of those categories have a harder time convincing network executives to put their vision on the air because there's no data to back the idea that those people will come out to watch themselves. Nielsen has taken steps in recent years to correct their data inaccuracies, by beginning to pull in information from Hulu and count DVR viewings, but they also admit that they have a lot to do to fix their mechanisms. Does this mean that we can blame the continued presence of Dads on our television sets on Nielsen? Unfortunately, no: luckily for Fox, the onus is still on them, for demonstrating that they'll give endless chances to uncreative people, despite the fact that there are so many more deserving ones out there.


Nielsen method for TV ratings missing minorities, young people [Poynter]

[Center for Research Excellence]

Image via Fox



A few things to note here (from someone who works in this industry):

-The CRE study is about the diary method only, and the reason why minorities and other groups are underrepresented is because they are less likely to be compliant. This is probably not as much of an issue with peoplemeters since they require users to do less - typically they will measure when the TV is tuned to a specific show regardless of viewer action, but might be less accurate if you don't log in so they know who exactly is in the room.

Otherwise, both the diary and peoplemeter populations are balanced to match the general population so they should not be underrepresenting groups from a composition standpoint. Nielsen spends a LOT of money and effort trying to get hard-to-reach demographics on their panels, and are also audited to ensure they are appropriately balanced. This can however introduce other biases - for example, hard to reach demos could get significantly higher incentives to stay on panels.

-For nationally broadcast network shows, the 21,000 peoplemeter households will probably be enough for most networks as compared to diary entries, which are significantly time-delayed but are essential for local markets. (Whether or not 21,000 is really still enough for a US ratings currency is a separate and hotly contested issue.) In addition to the Nielsen data, which provides demographics, networks can also use set-top box data from a variety of different providers for the purposes of just understanding whether a show is being watched. So I think for shows like Dads, the diary bias isn't what's keeping it on the air.

-Instead, some networks have gotten slower to pull the trigger because they think timeshifting might increase overall audience to a point where a show becomes viable (this helped Smash earn a renewal last year). Keeping Seth MacFarlane happy is another business consideration that helps out Dads a bit; and finally, some networks may also be figuring in social media interest/PR "buzz" in their cancellation decisions (unwisely in most peoples' opinions, but perhaps they are).