Inside Performance Advertising with Jason Fairchild delivers unfiltered insights, strategic perspective, and hard truths from inside the evolving world of adtech.
The industry is confused.
There’s a lot of hype right now about AI and creative. Amazon just announced an AI chatbot-style creative assistant. Comcast is launching a new tool to generate TV ads using AI. Magnite bought the CTV creative automation startup Streamr.
And to be fair, a lot of this excitement is justified. Lowering barriers to entry is critical to expanding TV to performance marketers and SMBs. Creative has long been a major hurdle. Making it easier, even free, to create a TV ad is a step in the right direction.
But the industry is missing the bigger picture.
Creating a TV commercial with AI is very different from using AI to make that ad spend perform. That’s the actual challenge and the biggest opportunity.
Let’s say I’m a broadcaster, and I roll out a slick new AI tool that generates TV ads for free. Great. Now the advertiser has a commercial. But what happens next?
They still need to spend. And more importantly, they still need that spend to perform. That’s where the conversation should be focused.
Search and social advertisers don’t care that a text ad is free to create. They care that $1 in drives $1.50 out. It’s about ROI, not production.
That mindset has created $200B+ in outcome-based spend. And yet, when it comes to CTV, much of the industry is still stuck in a reach-and-frequency mindset, mistaking the ability to produce an ad for the ability to make advertising work.
In short, spending money on CTV is easy. Making it perform like search and social is hard.
It requires more than creative. It requires optimization technology of the kind Meta and Google have built over decades.
That’s the core of performance TV: real-time optimization engines powered by feedback loops across creative, bidstream data, household ID, and declared outcomes like ROAS or CPA.
Let’s not undersell creative. It’s one of the many levers of performance. But it’s not the engine. We’ve run regressive analyses on thousands of creatives. For example, when the brand is introduced verbally within the first 1-2 seconds, there’s a 28% lift in performance. That’s useful. But it’s only actionable if it’s part of a larger closed-loop system that can ingest those signals and optimize accordingly.
At tvScientific, we’ve built AI that treats creative as a dynamic input to performance. Our platform classifies creative metadata in real time, learns which elements drive outcomes, and then automatically adjusts campaigns. If a certain style of ad, voiceover timing, or visual element boosts performance, the system generates more of what works, transforms assets into TV-ready formats, and feeds those insights back into bidding and optimization. In other words: creative is analyzed, evolved, and optimized for performance at scale.
Standalone creative isn’t enough. A raw, static creative is just that — raw — unless it’s integrated into a performance engine that learns and evolves in a scientific, data-driven way.
The same goes for measurement. Traditional adtech platforms talk a lot about CTV measurement, but unless you’re tying impression exposure to advertiser-declared outcomes using actual conversion pixels (either ours or via trusted third-party partners like Rockerbox), the data is meaningless.
We can optimize against any signal, but the system only works when all the signals, including creative, are integrated into a real-time feedback loop.
That’s the essence of performance TV. And that’s what the current conversation about AI creative is missing.
Inside Performance Advertising with Jason Fairchild delivers unfiltered insights, strategic perspective, and hard truths from inside the evolving world of adtech—cutting through the noise to focus on what really drives outcomes. Subscribe here.
