Ineffective Nielsen Ratings Are Ruining TV and No One Cares

By Brian Byrd | Think Pieces | January 9, 2015 |

By Brian Byrd | Think Pieces | January 9, 2015 |

Earlier this week, HBO announced that it renewed the half-hour comedy Girls for a fifth season. The news wasn’t particularly surprising. HBO typically publicizes its pickup decisions before or within a few weeks of a show’s premiere, and Lena Dunham’s polarizing, semi-autobiographical (you can’t spell Girls without “IRL”) look at four sentient garbage piles generates enough think piece ROI alone to justify another 10 episodes. What was interesting, though, were the types of adjectives some writers used to describe the series.

Multiple headlines featured variations on the following: “The hit HBO show Girls was renewed for a fifth season today.” Hit show? Was season three about professional piñata floggers who moonlight as contract killers? Because that’s the only way a show whose third-season finale drew 670,000 viewers should include the word “hit.” For the year, Girls averaged 820,000 viewers with only a bit more than half — 430,000 — coming from the coveted 18-49 demographic. More people eat at Chick-fil-A on Sundays than watch Girls.

As fun as it may be to stomp all over Girls’ head, it isn’t 2004 anymore. Using first-run viewership numbers to definitively declare that a premium cable series either is or isn’t successful is like evaluating an army’s strength by counting its cavalry brigades. Technology, busier work schedules and the original programming explosion have radically altered consumer viewing habits. Yet the measuring sticks remain unchanged. Those 820,000 weekly Girls viewers? Their ranks swell to 4.1 million when those who watch on HBO Go, On Demand, DVR, or during alternate airings are brought into the fold. For some reason, most reporters, networks and advertisers pretend these eyeballs don’t exist. Can you imagine this taking place in another industry?

Pete: “How many apples did we sell today, Sam?”
Sam: “Only 50. Well, unless you count the 4,000 we sold online and the 1,000 Frank paid for but won’t pick up until Wednesday.”
Pete: “So 50 then.”
Sam: “But we received money from those other transactions and have 5,000 fewer apples on the truck.”
Pete: “Don’t care. The Internet is a myth, like 9/11. Have you ever seen Loose Change?”
(Pete is Pete Carroll, obviously)

How is it acceptable to not know how many people watch a television show? We live in a society that can forecast congressional elections with pinpoint accuracy weeks before polling places open and determine a centerfielder’s defensive prowess relative to his replacement. Yet Big Television can’t implement a standardized system that accurately measures how many people consume their products? Other entertainment mediums embrace consistent metrics. Worldwide box office. Album sales and digital downloads. Physical and e-book purchases. None provide a flawless fiscal picture — ticket receipts don’t account for VOD viewing or piracy; more listeners access music through streaming services than ever before and artists derive much of their income from tours — but for the most part these systems correctly quantify successes and failures.

Television is admittedly more complex. A dozen factors go into determining what constitutes a hit. Some are measurable — series accessibility, multi-platform viewership, viewer demographics, award nominations, critical consensus, home video sales. Others are more nebulous — audience engagement, fan advocacy, buzz.

Check out the chart below. It contains 12 well-known scripted television shows along with each one’s Metacritic average, average Nielsen Live Plus Same Day (Live+SD) total viewership, and Live+SD 18-49 viewership. The bubble on the left indicates where the show airs, broadcast or cable. Instead of disguising each series with a bland “Show A” designator, the titles have been replaced with Wu-Tang Clan member aliases. Everything is more interesting with the Wu involved. Outlander should keep that in mind and cast the GZA in a major role.

So, which series would you consider hits based only on the information included in the chart?
(click to embiggen)

Not such a simple exercise, is it? Here are a few more data points that might further muddy the water: the average cable television show drew 1.84 million viewers in 2014; broadcast series averaged 5.84 million. By that benchmark, only four of the 12 shows on the chart can be considered above-average ratings performers (and Frank Stoney barely cleared the bar). Every broadcast representative and half the cable offerings are well below average, including the two shows — Osiris and Bobby Boulders — with the best critical scores. Conversely, critics seem to enjoy watching the highest-rated option, Tony Starks, about as much as they like seeing an orangutan fist their grandfather.

From a strictly quantitative standpoint, only three of the options can rightfully be labeled hits, with a fourth in the conversation. Now let’s remove the Wu aliases.

Amazing how much adding a name can alter perception. Despite drawing less than half the average viewers as a network drama, can we really pretend Hannibal, one of the best reviewed series and most buzzed about series on television, is a flop? The Knick’s 18-49 numbers would barely pass a Breathalyzer, yet the show earned a major Golden Globe nomination and was named the AFI TV Program of the Year. Veep, which won three straight Outstanding Lead Actress in a Comedy Emmys and been nominated for an Outstanding Comedy Series Emmy every year it’s aired, isn’t a success because it pulls less than a million viewers each Sunday?

Advertisers profit most from this uncertainty. They know the metrics stink. But there’s no incentive for them to adopt more accurate numbers; Live+3 (live airing plus any viewers who watch over the next three days) still determines ad rates. Live+7 gets closer to the true data, but both still omit DVR and digital viewing. Savvy companies exploit these market inefficiencies by purchasing underpriced ad time on purportedly floundering shows like Parks and Rec, then reap the rewards when a million viewers watch their commercial weeks after the original air date. A Live+SD standard doesn’t just shortchange shows and contribute to cancellations. Networks leave money on the table every time they miscalculate an episode’s true audience.

After years ignoring the mounting problem, some outlets are finally fighting back. It’s no surprise that the same cable networks who cultivated audiences through investments in progressive content now spearhead the ratings reform movement. FX already uses Live+3 for series premieres. HBO announced late last year that they will abandon next-day ratings in 2015 and move to Live+7/HBO Go/HBO On Demand hybrid model. In a statement explaining the shift, the pay-cable giant acknowledged that “subscribers have available to them an array of entry points to watch our programming — HBO linear feeds, DVR, HBO On Demand and HBO Go — and a single airing is no longer representative of an HBO show’s true audience size. Today, it is common for final gross-audience figures to grow anywhere from five to 10 times viewership after an initial airing.” Five to 10 times the audience. Imagine if movie studios undercounted their ticket sales by a factor of 10.

Even the broadcast Brahmins acknowledge the deficiencies. NBC, which probably has more incentive than any network to embrace accurate viewer numbers, recently asked its research department to develop a metric that reflects 21st century television consumption habits and encompasses the total number of people watching their shows across all platforms. The resulting metric — called the Total Audience Measurement Index (TAMI) — provides a staggering glimpse into just how much certain shows benefit when ratings include viewers currently ignored by Nielsen.

Take Parks and Rec. The half-hour comedy pulls a 1.2 Live+SD rating in the 18-49 demo and a 1.7 in the Live+7. When the 37 percent of viewers who watch the show through unmeasured digital platforms get rolled up into the total, Parks and Rec’s 18-49 rating shoots to a 2.7. Broadcast 18-49 average is 1.53. You don’t have to stretch your imagination too far to envision how this data could help sales execs secure higher ad rates or blunt advertiser objections.

It’s difficult to determine which series would increase the most from a Live+7/digital blend or TAMI standard. Shows that skew younger and those that air on competitive nights would likely see the largest gains, but an improving housing market doesn’t just increase your home’s value. It means your next pad will cost more, too. An inclusive, uniform ratings model creates greater success thresholds. If Parks and Rec’s true rating is a 2.7, how much higher is the real television average?

While revised metrics can clarify the quantitative picture, determining a hit show will always involve significant subjectivity. Networks still struggle to monetize social media engagement and may never truly gauge its impact. Award nominations don’t guarantee an audience. Even if it did, success is relative anyway. Goals vary. Executives may tolerate paltry ratings for a buzzworthy show if it eventually establishes a beachhead from which the network can launch additional, more profitable projects. The Walking Dead doesn’t exist without Mad Men. Given that, it’s probably wise to focus less on whether a show merits hit status and more on figuring out how many people actually watch the damn thing.