CTV measurement: Challenges and best practices

It’s time for the seventh edition of our MadTech digest, an outlet where you can learn about trends and best practices in MarTech/AdTech software development.?

But before we begin, make sure you read our previous editions about the?most innovative platforms for retailers ?and?retail media advertising .

With viewers ditching from traditional pay TV in favor of streaming services, the value of connected TV as an advertising channel keeps growing. eMarketer projects that almost 70% of the US population (230 million) will use CTV in 2023.?

This year’s CTV ad spending is expected to increase by 21.2% and reach $25,09 billion, surpassing spending on retail media networks (19.7%), another hot thing in digital advertising. As Netflix and Disney+ opened a new revenue stream with ad-supported tiers, ad inventory has grown in volume.

However, connected TV has its downsides. Accurate measurement and attribution remain the focus of CTV market players and the concern for media buyers.?

In this newsletter, we will learn about the connection between market fragmentation and measurement and attribution issues and discuss how businesses can get more insights into campaign performance.

If you want to learn more on the topic, read our comprehensive article on CTV measurement .

Data and device fragmentation of the CTV advertising market

The connected TV market players can be classified into four groups:

• Smart TVs devices with native OS (e.g., Samsung TV, LG TV, Sony, Vizio with built-in Chromecast)?
• Standalone streaming devices and media players (e.g., Amazon Fire TV, Apple TV, Roku, Chromecast)?
• Over-the-top (OTT) video-streaming services (e.g., AT&T TV, Rakuten TV, HBO Max, Hulu, Netflix, Disney+, Amazon Prime, etc.)
• Content distribution platforms (e.g., BitCentral, Amagi, Castify.ai, Viaccess-Orca, etc.)?

Roku’s recent announcement on making their own smart TVs and Amazon’s self-branded Smart TVs , available since late 2021, illustrate how dynamic and complex the CTV market is. It comprises numerous devices, media players, services, and video content distributors, each following its approach to audience measurement, attribution, and reporting. Since advertisers never focus on running campaigns in a single environment, ad platforms must pull data from multiple sources to give a complete picture of ad performance. The CTV market’s fragmentation, typical for new channels like in-game or DOOH , leads to considerable measurement challenges.

CTV measurement challenges

Lack of common identifiers. The digital advertising industry stepped away from third-party cookies to target users and establish attribution and works on replacing them with cookieless tracking and shared user ID solutions .?

Similarly, the CTV environment needs to devise cross-platform identifiers, as the IP-based identity approach doesn’t allow marketers to accurately attribute ad views and conversions to particular individuals or devices. Many household members watch content from the same streaming accounts but use different devices (i.e., the IP address changes, but the user remains the same or vice versa).?

Moreover, the use of IP addresses in the future is under threat because of privacy concerns: They are considered as PII under the California Consumer Privacy Act (CCPA) and GPPR. In particular, CCPA regards an IP as personal information if a third party, like an internet service provider (ISP), can link it with an individual or household using additional data (a phone number, physical address, name, etc.). GDPR treats static IP addresses as personally identifiable information; dynamic IPs can be PII if the ISP can identify users.?

In response to changes in the privacy landscape, Google introduced IP Protection (previously called Gnatcatcher), which proposes to avoid sharing users’ IP addresses with third parties.??

Numerous measurement methodologies. A multitude of measurement metrics is another consequence of market fragmentation.?

DSPs and SSPs providers use third-party measurement solutions because they need to provide buyers with specialized data on ad viewership, which CTV platforms don’t share. Also, big market players (Google, Samsung, Roku, and Amazon) don’t fully disclose how they measure audiences. That way, ad measurement tools give additional “points of view” on ad performance.

Fragmentation also exists on the AdTech level: Advertisers may have direct deals with CTV/OTT platforms or use ad networks to serve ads programmatically. The need to cross-check analytics results data from multiple sources complicates measurement.

Complex device identification. Since most platforms use IP addresses to identify users, it’s hard to get insights into ad viewership on an individual level: One person or several people can watch content on a smart TV.

One of the solutions would be implementing WURFL . The technology allows for identifying a device person uses to see an ad and extracting its characteristics (model, OS, browser, screen resolution and width).?

Cross-media measurement. Consumers watch streaming content however they please, so they use different devices, platforms, and services. As a result, data related to their activity is distributed across multiple systems. What is worse, most of the CTV market players don’t exchange audience data with each other. That’s why brands struggle to define a single customer across all devices to be consistent in their messaging, using CTV to raise awareness about a product or service in combination with other channels (e.g., in-app or web) suited for lower-funnel marketing.

CTV ad fraud. Complex attribution and the lack of universal standards for measuring ads, common user identifiers, or straightforward device identification approaches create loopholes for fraud. Faking ad impressions for CTV is easier than for desktop or mobile devices, as verified data is hard to produce (although sophisticated mechanisms of fraud detection might help).

Invalid traffic is a persistent issue for the CTV environment. It leads to wasted ad spend or lower return on ad spend for media buyers and decreased value and reputation of publishers’ inventory. The global invalid traffic rate for open programmatic ads on CTV devices grew to 21% in Q1 2023 from nearly 18% during Q3 and Q4 2022, according to Pixalate’s report .

Best practices of CTV ad measurement

Employ a hybrid approach to cross-channel attribution?

User attribution is complex because access to audience data is constrained. However, methods to identify users and monitor their interactions with content already exist.

IAB Europe recommends hybrid measurement approaches that include:

• Automatic content recognition (ACR) methods like audio fingerprinting or watermarking ?
• Passive panel metering technologies such as people meters or portable people meters
• Digital metering using linked mobile devices or home router-level meters
• Third- or first-party census feeds

Combining these signals makes it possible to minimize de-duplication and better distinguish between linear TV, broadcast video on demand (BVOD), or CTV app feeds at household and individual levels.

Determine the optimal approach to audience measurement

Audience measurement in cookieless environments like CTV is complex but possible. According to the set of standards and methods by the Media Rating Council (MRC), two main approaches are:

• Pixel-based technology to track an impression, capture video start and completion data, and detect and report invalid traffic.
• Embedded SDK or client-side measurement code for cross-channel measurement (e.g. Open Measurement SDK by IAB Tech Lab. The SDK enables publishers to offer advertisers to use their preferred third-party ad verification providers.)

Market players can use automated content recognition (ACR) technology on top of these approaches to enhance audience measurement. ACR identifies video and audio by capturing signals from playing content and cross-referencing them with a publisher’s database for matching.

Based on matches, AdTech platforms access ACR data for targeting, measurement, and attribution. Next, it can be cross-validated with passive or digital metering to improve accuracy further.

Choose the best ways to report on CTV performance?

Advertisers can use standard performance metrics for tracking CTV ads, such as quartile rates, completion and quartile rates, ad viewability, or CPM. However, these metrics are not sufficient to accurately analyze TV ad performance. Delivering ads and tracking impressions after TV sets are turned off is one such viewability measurement issue. In 2022, nearly 8–10% of CTV ads were served on dead screens; the problem costs marketers about $1 billion in ad waste yearly.

Organizations have been addressing viewability measurement challenges. The latest release of OM SDK (1.4 ) comes with CTV-specific signals like display connection status and last activity. The former allows for learning that the TV display is off while an application may still be running, and the latter indicates that a user is still watching content.

Furthermore, in February 2023, DoubleVerify released a new solution that “helps advertisers understand if ads play even after a TV is turned off.”

IAB Europe notes that cost-per-completed viewable view (CPCVV) is a more efficient performance metric than CPM or viewability, given the high viewability of CTV ads.

Final words

Connected TV is gaining traction, with viewers increasingly cutting the cord. Media buyers include CTV in their channel mix to expand their campaign reach. At the same time, device manufacturers, OTT streaming services, and distribution platforms monetize their audience data for the influx of ad revenue. Trade Desk’s and Magnite’s pivot toward CTV shows that AdTech businesses see the channel potential. Moreover, AppLovin added CTV inventory to the AppDiscovery mobile user acquisition platform and started charging advertisers based on app installs instead of impressions. The company’s decision changes media buyers’ perception of CTV: It’s no longer a brand awareness channel but a performance one.

The CTV market focuses on tackling ad measurement and attribution issues to ensure stable growth and maintain the businesses’ confidence that it’s the ad channel worth investing in. There is already some progress, but when do you think CTV will achieve this goal?

If you’d like to get a free consultation on building your own CTV ad platform or adding new functionality to your solution, contact the Xenoss team at [email protected] or via the form .?