Why is it So Hard to Blend Metrics Across CTV, Addressable TV and Linear TV?

As a marketer, you likely know that the wonderful thing about the huge explosion in ad-supported streaming content is that all the numbers line up perfectly and all metrics are straightforward, just like the web.

OK, so maybe things aren’t quite that simple.

Rather, as TV viewership shifts rapidly to become more streaming-centric, TV advertisers find themselves in a state of transition – colored by lots of confusion. The industry standard, Nielsen, lost accreditation this year as well, which only added to mounting turmoil for marketers. For as big as CTV has become over the last few years, linear TV and addressable TV still remain viable channels, and likely will for some time.

That means we’re stuck trying to make sense of a hodgepodge of metrics for very different mediums with very different natures and use cases, all under the banner of “TV.”

At the same time, the pool of TV advertisers is quickly broadening, from classic branding companies to digital natives, with vastly different goals and points of view.

Right now, there is no perfect formula for solving TV’s growing math problem. From our perspective, we find it useful to break down what we’re trying to measure with each set of numbers and what each brand’s ultimate goals are.

For starters, we think it’s important to make a distinction upfront. While there are all sorts of TV metrics out there, there are really two categories:

1. Campaign Delivery
2. Campaign Measurement

Let’s try to break each down and get to the root of what brands are trying to do when they look to combine data sets.

Delivery: What Did I Buy, and What Did I Actually Get?

As you probably know, the currency employed for different flavors of TV varies wildly:

• Linear TV buyers typically care about how many people in a given demographic are watching shows at the same time. But the way they actually purchase TV ads is via gross rating points (GRPs)
• Addressable TV buyers traditionally purchase ads that will only run in specific households at specific moments through cable set-top boxes – such as families with new babies at home
• Digital video campaigns typically traffic in views or impressions, and CTV has generally followed suit

Ideally, we could all just crunch each set of numbers and translate them into a common impression, making everything apples to apples. However, given that we are dealing with different bases, samples, panels and denominators, this isn’t easy. And keep in mind that many brands and agencies (though not all) are trying to evaluate all videos together – not just TV but YouTube, Twitch, Snap, etc. All of these platforms generate different numbers and have different definitions for views and view-through rates.

It’s quite a puzzle. Let’s call it the TRP Gap. In our experience, the best way to focus this exercise is to zero in on two key questions:

• How many people did I reach?
• Am I able to identify how often I’m reaching the same people on different platforms, and can I take measures to de-duplicate my campaigns?

Of course, each campaign is different and each has its own degree of complexity, but you can boil things down to: how many people in my target audience am I reaching and how often. Finding that valuable source of truth should allow a marketer to take step 2: measurement.

Measurement: So What Actually Happened as a Result of Running the Campaign?

Historically, when evaluating TV campaigns after the fact, most marketers start with cost. As in: What did I pay to reach my target audience? This CPM-based approach made sense when most TV advertisers were focused on awareness. Plus, with traditional linear TV, that was basically a brand’s only option beyond making broad assumptions about impact (sales seemed to go up when our brand ran x amount of TV GRPs).

With addressable campaigns, brands have been able to make an attempt at better correlating ad exposure with actions – but there is still typically a large data lag depending on the product you are advertising and your sales cycle. For example, people don’t run out to a car dealership right after seeing a car ad, even if they are in the market.

Digital media, of course, offers varying levels of more trackable and immediate response. The good news is CTV incorporates many of digital media’s attributes. For instance, it allows brands to have a very good sense of when an ad actually reached a user or household. The bad news: follow-through is harder to track, for example, and people don’t typically click on ads via their remotes.

Given this dynamic, many brands are finding the best way to connect TV ad exposure is through a device graph that allows for cross-platform and cross-device measurement.

Connecting the Linear, Addressable and CTV ‘Dots’

For this to work to its full potential – that is, if brands want to be able to figure out the contribution of linear, addressable and CTV toward driving specific outcomes – they need some sort of universal identifier. Indeed, the best way to take that unduplicated reach number I emphasized earlier and link it to performance data across multiple platforms is to employ a common identifier.

There are numerous attempts in the market currently aimed at creating a standard cross-platform identifier, be it email-based IDs or IDs that are unique to media companies, technology giants or data connectivity platforms, which pool together data from across the industry.

Another way to do this, as we do at Viant, is via household data. This takes into account household purchase decisions and the influence advertising has on a home, rather than trying to tie 1:1 exposures to conversions. As an example that many of you might be familiar with: my son saw an ad for his favorite marshmallow cereal and begged me to buy it. I didn’t see the ad personally, but I did put it on my grocery list.

Over time, I believe that brands and agencies will continue to consolidate their video buying into omnichannel platforms, as more TV is consumed through CTV and as more companies are able to answer the question, who did I reach and how often did I reach them? The key, no matter what, will remain understanding one’s goals and measuring for the right outcomes.