How To Accurately Measure CTV Campaign Results

The statement "I know half my advertising is working; I just don't know which half" takes on new significance for Connected TV advertisers.  

Because it’s possible no half is working the way you’d like it to. But you may not know it until the business results (or lack thereof) roll in. 

Accurate measurement has always been a challenge for TV advertisers. Streaming and CTV was supposed to be the solution—greater visibility into the exact impact of every impression. Digital-like accountability for the first time. 

But many CTV advertisers are finding that’s not the case. Or that the “digital-like” attribution models are failing to shed light on the bigger performance picture. 

 

What are the biggest CTV measurement challenges?

Audiences are more fragmented than ever. 

There are dozens of streaming platforms and seemingly infinite ways for streaming viewers to access content—on their phones, tablets, smart TVs. Keeping track of who’s watching what (and when and where) has never been so complex. In fact, a 2022 survey of US marketers listed viewership fragmentation as the top challenge facing converged TV advertisers. 

 

There’s no universal currency for streaming measurement. 

Cross-platform CTV measurement is made even more difficult by the lack of consistency and communication between publishers. Each publisher can use their own method for counting impressions. So a marketer trying to review their campaign’s reach on Hulu vs Pluto TV could be comparing apples to oranges. Plus, impressions are reported differently between linear TV (refers to people) and streaming TV (refers to households). To compare the two types of TV, conversions must be made. The bad news is there’s no standard conversion rate just yet. 

The TV industry is making strides towards a comprehensive approach to measurement, but for now, with no universally accepted method for attribution, advertisers miss clarity around the effectiveness of their campaign, whether they’re reaching the right people, or whether they’re reaching anyone at all.  

After all, 8-10% of streaming impressions are delivered when a TV set is turned off. And publishers overcount impressions from 2.5%-15% across all CTV streaming activity.  

 

CTV ad fraud continues to rise. 

In 2022, CTV ad fraud grew faster than CTV viewership itself according to DoubleVerify. And the problem’s far from resolved today—creating additional attribution challenges. 

Fake CTV impressions skew the data marketers depend on to assess the performance of their campaigns. This not only misleads advertisers about the true reach and effectiveness of their efforts but also drains ad budgets on views that never reach potential customers.  

 

The idea that CTV measurement is ‘easy’ is a misconception.

The wave of digital-like capabilities streaming and CTV adopted made it seem like CTV was the solution to TV advertising being notoriously difficult to measure. But CTV attribution isn’t easy. In many ways, it's even harder than measuring linear TV. And that’s because the technology’s only partway there. 
 
Gaps between user and device still exist for CTV measurement in a way they don’t for true digital campaigns. For example, the most common way for someone to respond to a CTV ad is via their phone, which means the IP address being targeted won’t always match the one responding. The current solution is using device graphs, but these are highly inconsistent from company to company. 

 

 

How to improve your CTV campaign’s measurability. 

So, CTV isn't the easy answer to TV advertisers' measurement dilemma. But just like linear TV, there are steps advertisers can take to make their CTV campaigns more measurable and ultimately have greater confidence in results.

 

Start by defining what’s worth measuring.

Clear results usually start with clearly defined objectives. Which is why smart CTV campaign planning shares similarities with planning for a measurable linear TV campaign.  

First, establish a quantifiable primary goal to assess campaign performance—what is the most important outcome you’re looking to get from your CTV investment? This is your greatest source of truth, and for many performance marketers, it will be some form of ROI or ROAS.

Second, identifying supporting KPIs such as improving website traffic, conversion rates, or brand awareness is also helpful. These indicators should be customized to your brand and business, providing a focused approach for evaluating results after the test is complete. 

Finally, make sure to spend enough to drive a measurable change in your chosen CTV metrics. Because the worst possible outcome of any TV test is not knowing if it worked. 

 

Use multiple models to verify performance.

Trusting a single metric can be deceptive. Embrace a multifaceted approach to attribution by running multiple measurement models in parallel to gauge your streaming campaign’s true impact. 

Bottom-of-the-funnel results tend to show up first. So after launching your campaign, start by analyzing response through calls, texts, app downloads or web traffic. 

Not everyone acts immediately after seeing an ad on TV. Over the weeks following your campaign, track changes in website traffic composition—TV primarily drives paid search, direct and organic traffic. Conversion rates should also increase since TV drives an especially high-intent customer. 

In the long term, look for improvements in brand awareness, recall, and familiarity. And of course, revenue. 

 

Use IP tracking to benefit from CTV’s digital-like features.

IP addresses have plenty of pitfalls. But they do offer a whole new perspective from which to evaluate results. Once someone’s seen your ad, you can track their online behavior all the way from impression to conversion, providing unique insight into how viewers respond to your ad. So while IP tracking for CTV is far from perfect, streaming advertisers should absolutely include it in their mix of attribution models—while keeping limitations in mind. 

 

Confirm results with incrementality testing.

One of the clearest ways to determine the impact of any marketing campaign is by analyzing incremental changes. When setting up your streaming campaign, separate your audience into a test group that will be shown your ad and a holdout group that won’t see it. Once your campaign’s running, you can compare outcomes between the two groups. If the group that was shown your ad sees higher orders, new customers, and conversion rates, that’s a strong signal the campaign’s working. 

 

Use fraud prevention tools designed for CTV.

Third-party fraud prevention and detection tools can help reduce the impact fraud has on your campaign—giving you a better understanding of your campaign’s performance among real, engaged audiences. Top providers include DoubleVerify, Method Media Intelligence, and Integral Ad Science. 

 

 

Confidently measure CTV performance with Annika. 

There’s been so much promise around streaming and what it could mean for the future of TV advertising. How it could turn a traditionally challenging channel into a highly accountable, highly measurable one. So far, executing on that promise has been difficult.  

But now, we’re taking a step towards real performance-first Connected TV. With our media-buying AI, Annika, advertisers get access to multiple models running in parallel. That way they can feel truly confident in campaign results. 

Learn more about how we verify CTV results or connect with us to discuss how Connected TV could drive results for your business.