CTV Will Flip Last-Click Attribution On Its Head

On TV & Video” is a column exploring opportunities and challenges in advanced TV and video. 

Since the 1950s, marketers have instinctively bought into the power of TV advertising, despite the fact that we have never been able to effectively measure its impact.

This somewhat inconvenient truth, coupled with the high cost of TV advertising, has made TV the private playground of an elite group of national advertisers who can afford to spend big on awareness and image marketing. More than 60% of the $74 billion TV ad industry is driven by only 200 advertisers.

But the rise of connected TV (CTV) is changing all of this with huge implications for the entire media ecosystem, from consumers to marketers to broadcasters. The ability to measure CTV in the same way and with the same analytical rigor as search and social channels is for the first time opening up TV advertising to the 9 million advertisers that are active in search and social media marketing today.

Call it the Second TV Revolution. 

The shift in consumer behavior toward streaming services is well past the tipping point, so it’s a fait accompli that CTV will disrupt the entrenched TV advertising industrial complex. The only real question is how much and how fast.

A recent Samsung study across its active CTVs (by far the largest relevant panel in existence) revealed that more than 60% of all viewing time in the US is now spent on streaming services. A massive, active addressable audience combined with digital-like measurement and attribution means major change is coming for TV advertising.

We’re not just talking about converting the current $72 billion in linear spend to CTV – we’re talking about diverting a significant chunk of the $150 billion digital performance market to CTV as well.

In other words, we’re in the process of doubling the size of the TV ad market, and that’s a big deal.   

The great CTV attribution paradox

Marketers can now embrace CTV to build a scalable performance-marketing channel that delivers the efficiency and effectiveness of search and social, but has far more powerful equity-enhancing attributes, such as brand awareness and affinity.

But for marketers to succeed in CTV they must first grasp this important reality: Users cannot click on their TV screens.

While this may seem obvious, the implications for marketers are profound. Simply put, CTV advertising will drive customers to engage in second-screen behaviors, meaning the “last click” inspired by a TV ad will come from a channel other than TV, such as search, social or direct type-in traffic.

Understanding this paradigm – and developing a framework for valuing it – is essential for growth marketers if they expect to seize the opportunity of TV content delivered on increasingly smart, increasingly captivating TV screens.

The greatest attribution heist of all time?

Who benefits from a TV ad that drives a Google search or a Facebook click? Google and Facebook, of course. If I see an ad for Domino’s on TV and I’m in the mood for pizza, my next move will likely be to grab my phone, do a quick Google search for the nearest location, click on the link and place an order.

This behavior will be attributed to Google in the Domino’s marketing dashboard, while in fact the search and the pizza sale were 100% inspired by the TV ad.

This faulty attribution dynamic has played out for years, resulting in hundreds of billions of dollars in enterprise valuation being hijacked by last-click digital companies instead of TV companies.

It’s also resulted in marketers avoiding TV as a performance medium. Case in point, while TV is dominated by mere hundreds of advertisers, search and social have millions of advertisers, and that’s because it’s easier for marketers to measure.

So, what’s to be done?

Through next-generation CTV ad platforms, marketers can now understand the CTV exposure-to-outcome impact of TV ads on a household basis, where the outcome is a website visit, a purchase or an app download. And marketers can still use their analytics platform of choice – often Google or Facebook – to track the last click a user made before arriving at their website. 

Because TV doesn’t show up on the Google or Facebook analytics dashboard, marketers need to develop an attribution model that reflects the value of TV. This is complicated, but necessary and achievable.

Some sophisticated performance marketers develop recency attribution curves that apply credit to TV-correlated website traffic based on the time gap between the last TV ad impression delivered and the website traffic during that specific time window. But this approach usually only assigns TV credit within a 15-minute attribution window. Conversely, when you can see the data related to CTV ad exposure-to-outcome on a one-to-one basis, the data clearly shows that CTV ads drive second-screen actions for many days, with three days capturing 70% of the conversions.

Trust but verify

Ultimately, though, marketers need to develop their own approach to TV attribution based on the new data now available from CTV ad platforms.

While attribution models are necessary, the ultimate trust-but-verify proof of value will come from incrementality testing. Marketers should either execute periodic incrementality tests or maintain an “always-on” control group to compare against the CTV-exposed group in order to validate the incrementality of CTV advertising. This will help validate assumptions and confirm theories.

A new era of “performance TV” is upon us

TV remains a powerful viewing experience in consumer media, arguably the most important screen in most American homes. The mass migration from linear to CTV is remaking a vast media ecosystem. Precise, digital-like measurability will change the way we understand and value TV advertising and – hopefully – evolve and advance our crude understanding of last-click measurement models.  

Advertisers will achieve better outcomes via TV, thus attracting millions of businesses to the medium. TV content producers will capture more advertising dollars from new TV advertisers, which will fund more and better content development. Consumers, meanwhile, will receive an ever-improving TV experience with better content choices and more relevant advertisements.

It really will be the Second TV Revolution.

Follow tvScientific (@tv_scientific) and AdExchanger (@AdExchanger) on Twitter.

Enjoying this content?

Sign up to be an AdExchanger Member today and get unlimited access to articles like this, plus proprietary data and research, conference discounts, on-demand access to event content, and more!

Join Today!

2 Comments

  1. Harmen Westra

    Jason - This is nice, but you may want to check out iSpot.tv. They have been doing your " TV Performance" TV/CTV attribution for over 4 years already and have built the appropriate value exchange of TV vs CTV investment relative to business outcomes for many clients.

    Reply
  2. To say TV isn't measurable does a disservice to what is otherwise a very strong article. Marketing Mix Models grew in use by the likes of P&G and the other top 200 marketers precisely because they could measure TV's contribution. One of the problem with MMM is reach. It has to be high enough to register in these aggregate (top down) models. This means MMM isn't a very democratizing technology. It works for the big guys, but not the little guys. Part of what is exciting about digital advertising is it is incredibly democratizing. With CTV, it is possible to measure at the person level (or household), and therefore CTV is more measurable at smaller spend levels.

    Jason, you are right to encourage always on control groups and incrementally and to eschew recency tests. The mix of a good TV design of experiment with digital conversion will unlock the incremental contribution of TV and place attribution for a sale in the proper column. But, if marketers miss setting up a control group, they will blow their chance to truly understand the contribution of CTV to a sale, which leads to misallocating budget and driving down sales and profits. Re-read Jason's suggestions. He is spot on.

    Reply

Add a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>