TV Analytics 101

The TV analytics landscape has taken so many turns in the last few years that broadcast stations have been in a constant state of whiplash.

It seems like just a few years ago stations could rely on Nielsen for a straightforward understanding of viewership. Today, cable, satellite, DVR, OTT (over-the-top), and digital have introduced several levels of complexity to the once simple TV viewership data world.

With such a complex collection of data, stations might not know how to best use all the analytics available to them. This post gives a brief introduction to the topic of TV analytics, including some tips on how to take advantage of advances in TV data analytics.

Not all data is created equal

Broadcast stations rarely rely on just Nielsen data anymore. (If they were to, they would forfeit major insights about their content.) And while Nielsen often remains the trusted currency for advertisers, many stations also buy comScore data. In fact in a growing number of cases, stations have now dropped Nielsen for comScore in ad selling.

For web-based content, stations likely use Google Analytics or other paid web traffic measurement tools to track performance. A growing number of stations also now use Spark Station Analytics to track near real-time viewing patterns and competitive performance. And the list of data solutions could go on. Although, these four are the most commonly used among stations.

Any remotely accurate analytics source is better than none, but not all data is created equal. And while all TV data companies hope to offer accurate data around viewership numbers, they arrive at those numbers by using very different methods. Here’s a look into what the main four TV station analytics companies are measuring and how they measure it:

comScore

What they measure:
TV viewing habits and trends from households, primarily with satellite TV
How they measure it:
Gather viewing data from satellite companies’ set-top-boxes at individual homes and extrapolate total viewership
Pro: Data is collected from a relatively large sample of homes
Con: Slow process of getting the data in the hands of the stations

Nielsen

What they measure:
TV viewing habits and trends, which are summarized in Nielsen rating points
How they measure it:
Gather viewing data from a limited number of homes, via either paper digests or people meters, in a given DMA (these homes comprise what Nielsen calls a “panel”) and extrapolate a percentage share (rating), which is then sold to stations
Pro: The incumbent data source is recognized and generally trusted by advertisers
Con: Very slow process to get the data–days or weeks after a program runs; relatively small sample size

Google Analytics

What they measure:
Website views, clicks, and other interactions
How they measure it:
Code that tracks website visitors’ activity on the site
Pro: Very exact and immediate data, making it easy to understand which stories resonate based on web traffic data
Con: Strictly limited to web activity

Spark

What it measures:
Second-by-second TV viewing trends and habits, including competitor data
How it measures it:
Smart TVs send all viewing data from a given DMA to the cloud-based Spark platform accessible in near real-time by statons in the DMA
Pro: Not reliant upon paid TV subscriptions, panels, or lengthy collection and delivery processes
Con: Strictly limited to smart TVs

There’s the quick and dirty view into what types of data sources are out there. Your station might even have access to all four data sources. But the value of data doesn’t come from simply having it; the value comes from knowing how to use it to take your station to new heights. The next sections give tips on how to best use your station analytics.

Use multiple data sources to tell a more convincing story to buyers

Nielsen data has ruled station analytics for decades, largely because technology limited any other data collection. And still, sales teams often look to Nielsen ratings as the currency in ad selling.

Advanced TV data analytics options like comScore and Spark don’t have the legacy power that Nielsen has with advertisers. They’re relatively much younger in the TV ad economy. But stations that don’t use advanced analytics as part of their sales pitch are missing out on selling potential.

With Nielsen ratings, for example, you can sell a 10p ad on the understanding that a potential audience size and demographic (estimated through modeling once a quarter in many cases) will see it. But if you add comScore and Spark Station Analytics data to Nielsen’s you can show detailed historical and even real-time audience data to strengthen your story and raise the advertiser’s confidence in how their ad will perform.

The more credible data sources you can use to sell your ad inventory, the more confidently you can demand your desired price. Chances are, many stations could demand higher prices of advertisers if they used all their data sources to more completely understand the value of their inventory.

Instant insights vs. long-term trends

One report on viewership for one show on one day can tell a lot about the performance of that one moment on a channel. When it comes to making data-based decisions, though, there’s an important call to make in whether to look at instant insights or long-term trends.

Instant insights come from data that tell about a certain program on a certain day, or perhaps even a moment in that program. This kind of data is invaluable to a station. You can boost station morale by praising producers and reporters of successful stories.

You could also use the data to spot red flags. Perhaps one segment of a program lost 50 percent of the audience at one point. This poor performance could trigger you to look into why it failed, whether there’s a solution for the future, or even whether to consider cutting the segment. Just be careful to consider whether it’s the story itself that performs poorly or whether the story is simply aired at the wrong moment in the program.

Sweeping decisions around stories, programs, and ad placements should not be made based on instant insights. You would have to make a lot of hasty assumptions to cut a show, program, segment, reporter, etc., based on one isolated, instant insight. This is where the importance of long-term trends comes in.

A long-term trend is four or more weeks of data around a program. This kind of data is just as critical to a station as instant insights. After reviewing longer-term data on a program, or segment of a program that performs poorly you might consider making tweaks to it (what time it runs, where and how-long the ad spots go, etc.). If it continually falls short of expectations, even after some tweaks, then you might consider replacing the content with something else.

Once a station really starts to involve data in decision-making, it can start operating with more confidence. Gut decisions become the exception, not the rule, when you take advantage of the data available to your station.

The more advanced the TV analytics, the more relevant the station

As acknowledged at the start of the post, the last few years have been a wild ride for TV stations. The changes in TV analytics might seem overwhelming at first glance, but the advances in data measurement can actually make a news producer’s job (or a salesperson’s, or a research director’s, or a promo director’s job) easier and more successful. And as a whole, the stations that embrace new TV analytics technology will be the stations that stay relevant and experience the most consistent success as the TV landscape continues to evolve.

Sam Petersen
Sam Petersen
Content Marketing Manager
Sam is a conflicted marketer. He loves being creative, but also thrives on numbers and analytics. Sam has worked in several messaging-focused roles during his career, including journalism, PR, product marketing, and now, content marketing. When he's not stewing over his conflictual identity, he heads to the mountains to hike, ski, run, etc.