The Role AI Can Play in CTV Measurement
When you think of AI, what is the first thing that comes to mind? Maybe it’s the chatbots that you talk to on Ticketmaster. Maybe it’s the filters on TikTok. Regardless, your first thought probably isn’t AI as it relates to CTV and measurement. But maybe it should be.
At this pivotal injunction in the CTV measurement evolution timeline, AI and machine learning need to be two breakout stars. The CTV industry, as a whole, has been on a rapid growth trajectory for years, but its path to mainstream has inadvertently left a void in the measurement space. While innovations in content production and distribution abound, innovations in CTV measurement lack gusto. That’s about to change.
The reliance on legacy
Ever hear the phrase, “the devil you know is better than the devil you don’t”? Even though the advertising industry has evolved tremendously, it is resting on legacy methods for measurement and data processing that are simply not cutting it anymore with the speed at which the industry moves.
For decades, advertisers have relied on specific identifiers like an IP address to target and understand consumers on CTV. However, with the rise of privacy concerns online coupled with the growth and interest in using AI, there is an opportunity to rethink what the future looks like.
The benefits of AI in CTV
Moving further into the future, there is an expectation that one day, AI will simply be another tool in the adtech tech stack. It should be second nature to the industry, but we aren’t quite there yet.
The name of the game for advertisers and brands is audience targeting. AI has the capability to go through thousands of datasets on consumers and analyze that data. This can help advertisers determine the best ways to reach targeted audiences and better align ads with consumer wants and needs.
A benefit of AI in CTV is that it is going to simulate household-level models without actually needing households. AI will allow brands to produce household level models to reduce fraud, understand impact of ads, and make better programmatic decisions on IP address or alternative CTV identifiers. This process will only help brands and advertisers to be compliant with evolving privacy laws and it will drive more efficiency in meeting their ultimate goals.
Another benefit of AI is automation. Automation allows for advertisers to more quickly understand and implement feedback they are getting in real time. This will help them adjust campaigns based on what is resonating with each specific consumer, thus leading to better outcomes. Additionally, they can better analyze data once campaigns have ended to determine patterns within flights. The data shows what works and what doesn’t when it comes to the audiences advertisers are targeting. With AI, advertisers can almost predict what is going to perform well and invest more time and resources into those systems versus the ones that aren’t resonating.
But it all relies on accurate and clean data. When you have humans working closely with data, inputting dataset after dataset in a mundane, repetitive task, human errors occur. When you automate the process, you eliminate those possibilities for error and increase efficiency and speed in decision making. AI enables an improvement in data quality which ultimately leads to more accuracy in measurement. AI will need to train on data that you have on a subset of TV households connecting their exposures to attention/brand lift/offline outcomes. It is important that publishers/ad platforms develop privacy friendly solutions for understanding CTV performance if pixels cannot be placed, partnering with third party measurement providers where needed (clean rooms etc).
The CTV industry has expanded exponentially over the last few years. It has become a hotspot for advertisers but a platform that is increasingly difficult to measure. As ad dollars are pinched, it’s critical to prove their value. AI has exploded onto the scene recently and its implementation to nearly every industry will not slow down anytime soon. And while we haven’t seen an incredible increase in the use of AI in adtech and CTV, it’s only a matter of time.
[Editor's note: This is a contributed article from Cint. Streaming Media accepts vendor bylines based solely on their value to our readers.]