I spent over 8 years working in the production and marketing of local TV news. I certainly had my ups and my downs in that time but I thought I had a fairly good handle on the business after close to a decade in the industry. I was dead wrong. Yes, I knew the content side and the creative side. But it wasn’t until I made the move to sales that I really figured out just how upside down the local TV news business truly had become.
Working in television sales was awful. There are so many issues that are worth highlighting but I’ll focus on the main problem. I’m not even going to touch on how shitty it is that the lines of real news and sponsored content have been blurred to the degree to which they have. I think John Oliver did a masterful job of that already and I can’t possibly improve on it. Instead, I’ll just focus on the core business model.
The problems with the TV news selling process run far deeper than just sponsored content. Like many of the societal issues we are currently facing in America, to understand the genesis of TV’s problems we must first start with the currency.
Ratings
Currency is a word that is of considerable importance in the TV business. The word is defined as a medium of exchange. The most well known currency is the one that many people transact in on a daily basis, the US dollar. Because of the distribution I’ve chosen for these writings, If you’re reading this, you’re probably already well aware of the problems with USD as a currency. In the TV business, the currency is “ratings.” When you get to a certain age, the word “ratings” becomes common nomenclature though few actually know what it even means or how a rating point is calculated. That goes for both viewers and people actually in the business.
A rating point, simply, is the average percentage of the available audience that a program acquired for the duration of the telecast. For instance, the Super Bowl is famously the highest rated telecast on American TV year after year. But the ratings in each city vary based on how many TV households are in the market. When you hear about the ratings for big ticket telecasts like sporting events, you’re more than likely hearing the national ratings. Each localized market (DMA) has its own rating figure as well. But you have to remember, the rating is a percentage of the available audience. And that audience is not only bigger or smaller depending on which city the viewers are in, but also how many of the households in that city are considered TV households.
This is an important concept to understand because TV households make up the Universe Estimate (UE). And it is the UE that is ultimately the denominator in the ratings calculation. Let’s say there are two different cities that each have 1 million people living in the DMA. We’ll call them City A and City B. If Sunday Night Football gets a 15 rating in City A and a 15 rating in City B, it might be assumed that each city produced the same amount of viewers of the game. Chances are it didn’t. The reason it didn’t is because of the UE. If City A has 500,000 TV households and City B has 400,000 TV households, a 15 rating in City A produced an audience of 75,000 households while a 15 rating in City B only produced 60,000. I promise I’m getting to why this is a big deal. Stick with me, we’re almost there.
The reason this matters is because cord-cutting has accelerated the decline of TV households and the rise of broadband-only households or BBO households. These shifts in household designations have contributed to fluctuations in the UE for every DMA. As the UE decreases, the amount of households that a rating point represents decreases with it. This is the great lie of “ratings.” A 10 rating in 2015 is not the same apple as a 10 rating in 2020. It’s critical to understand this because it’s the reason why ratings as a currency should go away and be replaced by something like viewers or impressions. Why don’t we just include BBO households in the rating? Long story short, the ratings services are trying, but they’re not there yet.
Methodology
Historically, there has really only been one company tasked with measuring the ratings of TV programs. That company is Nielsen, though measurement methodologies have definitely improved through the years, the old way of assigning ratings was done through what were referred to as diaries, or survey books. These diaries were literally printed out books that were mailed to participating households. The people who received these books would fill out what programs they watched. They’d note when they watched the programs, how long they watched for, when they flipped the channel, what they flipped to, how long they watched after flipping; you get the idea. When the survey was finished, the books would be sent back to Nielsen, Nielsen would compile all of the data, and then after a period of time, report the findings. These findings? You guessed it, ratings.
It seems crazy now given how immediate and how targeted digital advertising is, but for most of the existence of TV measurement, ratings were only compiled in February, May, July, and November. This is why TV stations used to bring out the big guns in those key months. The networks would run important programming like season finales in May. Newsrooms would plan their big investigative stories for those months. These months were called “sweeps” periods.
Now, aside from the fact that twelve months of media buying were largely decided by just four months out of the year, the added element of absurdity in the ratings methodology was that the viewing logs were physically filled out by the participants with paper and writing utensils. It was expected by Nielsen and by the station groups subscribing to Nielsen’s service that these survey books were filled out every day for the entire month. If a person with a survey turned on the TV and started watching their favorite show, they were to write it down and log the activity at that time. If after 15 minutes they decided to change the channel, they were to log that too. While I’m sure there were some exceptions, most people were unlikely to fill out these diary books in this way. Instead opting for last minute guessing when they remembered they needed to fill them out at the end of the month.
The “winners” of the entire broadcast industry were determined by the results of these diaries. On both a local and national level. We’re talking about content decisions, talent decisions, and most importantly, spending decisions. All largely made based on the information in these survey books.
It should be mentioned that there was compensation to the viewer for participating in the diaries. It wasn’t much, but if you had nothing better to do and you needed the money, it was an easy way to get a couple extra bucks during the year. Now, anyone who knows anything about human behavior knows that this is an awful way to measure content consumption. People are lazy. They would forget they were supposed to do it. Then at the end of the month when they wanted to make sure they got the money, they’d panic-finish the survey book with as much as they could remember and just guess at the rest.
This obviously led to a lot of problems. Networks would often find that Nielsen credited certain programs with absolutely no viewing. Imagine, a massive tower broadcasting a free video signal to hundreds of thousands of homes and not a single viewer would register. Sound plausible? Of course not. Another problem would be Nielsen would credit other programs with far too much viewing. Early on as a consumer analyst, I learned that surveys were great at measuring preference or brand health but terrible at measuring true behavior. I touched on why that’s the case briefly in the last chapter. People just generally aren’t honest with themselves and might think they watch the news every day when they really only watch a couple times in a given week.
Thankfully, Nielsen and it’s younger competitor comScore don’t measure TV viewing in this way anymore. comScore actually never did. Nielsen has a handful of different methodologies that vary by market. As you might have guessed, these methodologies also have considerable flaws. While the days of paper survey books are gone, many larger market cities rely on a more electronic version of the viewing logs. When a Nielsen participant turns on the TV, they are to identify which person in the household is watching. Anytime a Nielsen participant household has a guest in the home, they are to log a visitor. This can potentially skew data if it isn’t logged accurately.
One recent example of this data becoming skewed is from a Nielsen home that was hosting a house guest for an extended period of time. The guest was incorrectly logging viewing data and Nielsen’s system ended up extrapolating dozens of people in one home, all watching the same crappy network television. Because of how the methodology works, a sample increase of that magnitude in one Nielsen home created a projected viewing level that was far beyond what it should have been if the guest’s viewing log had been measured accurately. The point is, even in the digital age, when you require active participation from the lab rat, the data can get fucked up pretty easily.
For now, it’s important to understand this background and what ratings have historically been because it helps grasp why buying and selling TV currently is still such a mess. Again, ratings are the underlying currency. And the best currency in broadcast television is still adults between the ages of 25 and 54 years old. Why? Because these are the people who we think are at the peak of their household spending. They’re moving up in the world professionally. They’re getting married. They’re having kids. They’re buying houses and cars. They’re acquiring wealth. They’re exactly who advertisers want to reach. This group of people, adults between 25 and 54 years old, are what is referred to in the business as “the demo.”
Targeting “the demo” is stupid
Don’t believe me? Let's dive in. Targeting the demo in 2021 is stupid because the demo is a hypothesis. This hypothesis was the best approach to buying and selling media before the internet. But now we have the internet and we’ve had it for a pretty long time. With the internet has come data. Big data. We don’t need to hypothesize anymore. We actually have a better system for buying and selling already.
Ratings used to be determined by survey books during the months of February, May, July, and November. They aren’t anymore. Nielsen now measures viewing through several different methodologies. None of them require survey books. And this measurement is happening all day every day, 12 months out of the year. There are meters, audio watermark readers, and something called return path data (RPD). RPD is essentially a coded log of what viewers are watching through their set top boxes. So if you’re in one of the roughly 75 million households in America that still has cable or satellite service, there’s a high probability that Nielsen and comScore know everything you’re watching.
Ready for the really scary part?
They don’t just have your set top box. They have your internet service provider. They have your search history.
They have you.
Who are “they?” They are big data firms. You can probably guess who they are. If you don’t know, Google it. Hey, I just mentioned one of them! There’s a reason why when you search for a shirt on Amazon and don’t buy it there’s a good chance you’ll keep getting reminded through some other platform that you should totally buy it. This is the digital age. Buying advertising is completely different than it was just 15 years ago and the smart companies are doing it right.
The smart companies know that you don’t have to buy an adult between the ages of 25 and 54 years old and hope they’re in the market for a new car. You can buy someone who has been poking around dealership websites or looking up car reviews online. You can geographically target people who have set foot in an auto repair shop and hit them with a new car advertisement on Facebook. There’s a decent chance this person is or could be a vehicle intender. It makes more sense to try to advertise to this person than a random 45-year-old if you’re truly just trying to sell cars.
Now to be fair to television, the speed of processing this data and churning out an intender is a little bit slower. But what’s important to understand is the internet and updated measurement methodologies have made it so this targeting is still possible even on linear broadcast television. comScore, for instance, takes your RPD data from your set top box and matches it with your address and your internet provider. From there, they can see which households that watch The Masked Singer on Fox were also searching for Toyotas online recently. With all of this intention targeting capability on TV, guess what TV agency buyers and local TV sellers are still using as “currency?”
Ratings for “the demo.”
It’s pretty dumb and nobody on either side of the table is willing to change the currency because it’s hard. Advertising agencies and media sellers have been incredibly slow to adapt. If the flaw in this process hasn't quite landed yet, I’ll give you an analogy. Imagine you’re a manager of a pizza place and you want to hand out coupons to people in your town to drive traffic to your store. I’m running a consumer marketing event and I come up to you and I say, “I have two rooms filled with people and you only have time to enter one room. Room A has 30 people who are hungry and love pizza. Room B has 100 people and all we know is how old they are. Which room do you want to choose?”
Duh, Room A.
Now let’s pretend you’re not passing out the coupons. You pay a guy named Bill to pass out the coupons for you and Bill always picks Room B. Why? Because Bill also passes out coupons for other businesses in town and Room B has more people. Ultimately, there will be people in Room B that take some of the pizza coupons. There will be people that probably take some of Bill’s other coupons too. It was easier for Bill to pick Room B because he probably satisfied each of his clients enough to make entering the room on their behalf worthwhile for all of them. But we know you had a room that was a better fit because all 30 of the people in the room were the people you wanted to reach. Room B, though it had 100 people, may have only had 15 who were hungry and wanted pizza. TV buyers are still largely buying Room B and they shouldn’t be because one size fits all doesn’t work in advertising.
Points
Now that we’ve established what ratings are and why demographics are a bad way to buy and sell advertising, it’s time to really hammer home the idiocy of the process. The ratings are called “points” by ad buyers. If a local newscast averages a 12 rating, buying one spot in that program would constitute buying 12 “points.” Points are the real currency and this is where years of egregious complacency from TV advertising sellers has irreparably hindered the business model.
Once upon a time, everyone watched TV and it was easy to sell the broad reach of broadcast television. Since prime time ratings were strong with younger viewers and the ad spend in the industry was robust, TV ad sellers allowed TV ad buyers risk-free campaigns. What’s a risk-free campaign?
When an ad buyer calls up a local TV station rep and starts the negotiation process, they’re not negotiating how many commercials a client will get or how much money will be spent. They’re negotiating cost per point (CPP). How much does one rating point cost? Like almost anything, when you buy in bulk you get better pricing and TV is no different. When a buyer calls up a seller and says “I have $10k from the local Chevy dealer but I need a $16 CPP, take it or leave it,” the buyer is actually getting a risk free buy of 625 points if the seller takes the deal.
The purchased points are then delivered through commercials in various programs. For instance, if the buyer in our hypothetical deal above gets placed in an NFL game that delivers a 20 rating (20 points), the rest of the campaign only has to deliver 605 more points to fulfill the obligations of the agreement. If the finished campaign fails to deliver the 625 points and only delivers 500 points after the schedule has run, the seller has to make the buyer whole on the under-delivery by making up the rest of the points in a second campaign at no cost to the buyer. This is a “risk free” buy and it’s bad for a variety of reasons.
It’s obviously bad for the TV station because there’s no upside to the buy. If the campaign actually over-delivers and the buyer gets 700 points, there isn’t any additional money that gets ponied up by the buyer. If the campaign only delivers 500 points, the seller has to offer up free inventory to keep the buyer happy. This could put the seller in a supply crunch situation where they may end up crediting the missed points and giving money back. This setup will ultimately end badly for the buyers too because they never lose so they never have to improve their internal protocols either.
This agency edge is why buyers have no interest in changing the process. The sellers aren’t really in a position to demand changes to the process because there is too much competition for advertising revenue from the different channels and platforms. If TV sellers want the money, they have to take the deals and just hope they deliver the points. But I can’t stress enough, this one-sided relationship is from years of complacency in the TV industry on the sell side. TV sellers used to have something that every brand needed. Because of this, the money came easy because the audience was a slam dunk. Now, the audience is leaving and the money isn’t easy but the rules are still the same. TV sellers keep taking the no-upside business because they kind of have to.
The biggest problem though is this points structure doesn’t necessarily benefit the actual client. Because remember, the ad buyer is probably an agency buying points on behalf of a small business. The truth is, small business owners don’t give a shit about points. Small business owners want conversions. And conversions are ultimately the metric that prove campaign effectiveness. A campaign that delivers 700 points might not actually be as effective as one that delivers 500 points because points don’t factor in targeting efficiency and not every client needs the same thing. This is why buying an age demographic doesn’t make much sense anymore.
Fixing this mess would be simple but it requires an important psychological tweak; the agencies that run marketing for small businesses must start viewing advertising as an investment as opposed to an expense. They must view themselves as fiduciaries rather than as commodity traders.
In the end
To summarize this chapter as simply as I can; TV buyers and sellers haven’t changed the way they do business in decades. Despite the emergence of better data and better metrics to gauge the success of a campaign, nobody can get out of their own way. Buyers pretend they want sellers to improve the system (they really don’t because then they lose their edge). Sellers want buyers to improve the system. Neither want to actually take the necessary steps to change because change is hard and teaching old dogs new tricks isn’t likely. When you factor in that agencies charge clients for their services, it becomes even more evident that the client is ultimately the biggest loser in this.
Complacency and easy money. Years from now when we look back at why local TV news died, many will pontificate the reason for the demise. Biased journalists and “fake news” will likely get some blame. But that will just be low hanging fruit from political grandstanders. The biggest reason that will probably go unmentioned is the revenue model just completely broke down. And this is just the advertising sales revenue stream. Wait until we dive into distribution in Chapter 5.
