DigitalEssaysFeaturedInsightsMediaPartner content

TV attribution drives better understanding

By Youmna Borghol, Chief Data Officer, Choueiri Group (partner content)

By Youmna Borghol, Chief Data Officer, Choueiri Group

There is no question that TV reigns supreme when it comes to building brands. Marketers have turned to television for years due to its mass reach and scale, and its ability to drive awareness, social engagement and influence. But while TV has traditionally enabled viewing reach and scale, it lacked the measurability and accountability of digital advertising, particularly with campaigns that demanded direct response.

Brands frequently noticed an uplift in sales after launching a TV campaign (and a drop when they stopped), but correlation alone doesn’t cut it. It’s easy to confuse correlation with causation when you’re trying to measure TV. With the lack of direct, clear, down-the-funnel attribution models for television, measuring its impact on sales has historically been restrictive. Until now, that is. Enter TV attribution

With the evolution of TV attribution over the last few years, measuring a sale that is directly linked to TV has become easy. Brands can now tie the spot airing to an immediate action and measure the online impact of TV ads. TV attribution is giving brands the ability to better understand how their TV advertising is driving business outcomes by measuring responses and effectiveness in the same way digital media has for years. Marketers now can make the connection between traditional audience metrics (GRPs, CPMs and ratings) and brand-specific, performance-based metrics.

TV drives digital responses

The rise of second-screen devices has made TV advertising more powerful and measurable. Today, consumers can directly respond with measureable actions. All it takes is the touch of a smartphone. Even with all the distractions of today’s multimedia, content-driven world, we’re consistently seeing that traditional TV ads drive strong digital responses on a number of different channels (search, site traffic, app activity, social, etc.).

As consumers jump online after seeing TV ads, advertisers are using traditional TV as a powerful tool to drive audiences directly into the customer journey via digital. They are then using TV attribution to measure digital responses and evaluate how effective their TV ads are, what’s working across their TV marketing plans and what needs to change.

How TV attribution works

TV attribution allows brands to precisely track, measure and optimise TV responses.

It tracks a brand’s real-time response data (website, call or app activity) and overlays TV exposure data, to capture any incremental responses above baseline levels (that a brand would experience without TV ads).

The online impact of the TV ads is broken down for analysis by networks, programmes, genres, dayparts, times, audience segments and creative, showing what’s working and what’s not across TV campaigns. These insights enable brands to identify media-buying opportunities that will drive an optimised response, informing them how and where to adjust TV ad spend.

Getting TV attribution right

A key challenge for marketers is to do TV attribution the right way.

While there are a number of modern TV attribution solutions available, applying the right methodology is what makes the information it produces actionable. In my opinion, it is the model that is the deal breaker. If the calculation for measuring TV audiences is flawed, it will have profound implications, as the whole objective is to optimise TV via the data produced. For accurate attribution, brands need to understand how the algorithm works. There are three key points to understanding any TV attribution solution:

Baseline Calculation: In a simple world we would calculate a baseline level of website traffic, identify a spot’s start time and measure the uplift from the baseline for a period of time after the spot’s broadcast. Unfortunately, no baseline will ever be perfect. Some of the web visits may have been caused by TV, but others will just be random – the attribution model needs to be able to distinguish between random and genuine TV responses.

Attribution for overlapping TV spots: The attribution model will need to define the rules that determine how response is apportioned across competing spots. Understanding how the algorithm builds response curves and handles overlapping TV spots is crucial to ensuring that very small channels do not get over-estimated in attribution.

Data granularity: An attribution model is only as good as the data that supports it. Only by looking at data at per-second granularity is it possible to identify and isolate the subtle fluctuations that occur on an advertiser’s website or app when spots are broadcast very close together. The granularity means we do not over-attribute by misinterpreting noise for response.

Short vs. long-term impact

TV attribution is a short-term measurement solution focused on the immediate impact of TV ads. Of course, TV advertising continues to be impactful long after the first airing. As an industry, we need to be careful about how we use TV attribution so we don’t end up measuring TV ads with the equivalent of click-through rates. Let’s not fall into the same trap that we did with digital and focus greatly on a metric just because it’s easy to measure.

Ad spend coming back to TV

In 2018, more and more brands came back to or increased their investments in TV. As digitally-focused brands are approaching an inflection point when it comes to growth, TV could prove to be the biggest opportunity for their continued expansion. By combining the technologies that power our online world with the wide-range appeal of TV, broadcasters can not only stay relevant but also stand to thrive.

Comments