AI Search Isn’t a Threat to TV. It’s the Reason TV Matters More.

AI search is changing how people discover brands, and that makes TV attribution more important than it has ever been. Here’s why the shift favors TV and CTV, and how to measure it honestly.

Everyone treats AI search as bad news for advertisers. But if you spend on TV, it may turn out to be good news, because TV and CTV become the biggest winners.

The usual worry is whether ChatGPT will replace Google. But that’s the wrong question, because search isn’t one thing. It’s three.

AI is taking over discovery. Ask ChatGPT for the best life insurance and it just answers, no ten blue links, no ad auction to win. It might affect brand research. But it barely touches navigation. And that gap is the whole point.

For twenty years, the easiest way to grow a brand was to catch people who already wanted something. Someone searches for a product, your brand shows up through paid search, and you win the sale. But when someone is still discovering, before they know a brand name, AI now makes the recommendation before you get a chance to show up. Good advertising doesn’t create searches, it creates preference. It makes people want you before they ever search, so they come to you by name. That’s what TV and CTV do well.

Category demand vs brand demand

As a brand, you’re a step ahead once you understand this. Compare two people:

Person A asks an AI, “What’s the best car insurance?” That’s up for grabs. The AI or a competitor can point them anywhere.

Person B types “Geico” into their browser. That decision is already made. No algorithm is going to talk them out of it.

The first is category demand. The second is brand demand. Brand demand is much harder for anything to intercept. The question is how someone gets from A to B, and that’s exactly what demand creation does.

Why this favors TV and CTV

TV was never good at catching demand, but it shines at creating it. Its job isn’t to show up when someone’s already shopping. Its job is to make people want your specific product before they ever pick up a device.

Strong TV moves people down the funnel ahead of the search:

• From “I should shop around for car insurance” to “I want to check Geico.”
• From “I should support a children’s hospital” to “I want to give to St. Jude.”
• From “I need to plan for retirement” to “I want to call Fidelity.”

By then the person isn’t discovering a category. They’re heading somewhere specific.

CTV takes this further. It does the same demand-creation job as traditional TV, but it’s addressable and measurable in a way linear never was. You can target the households you want, and you can connect the ad to what happened after. So you’re not just building preference at scale. You can actually see it work. That combination, creating demand and proving it, is exactly what matters more as AI eats the easy capture at the bottom of the funnel.

To be fair, TV is rarely the only reason someone lands on a brand. Podcasts, word of mouth, influencers, and years of other marketing all feed into it. TV is one strong driver of preference, not the only one. Which is exactly why measuring it correctly matters.

Why TV attribution has to change

In all this, the real risk for advertisers isn’t losing traffic. It’s judging success with the wrong yardstick.

TV attribution is how you connect a TV or CTV ad to the business results that follow it. For a long time, people used branded search volume as that link, to “prove” TV worked. TV airs, searches go up, call it a win. But search was never the goal. It was just a footprint. The goals were always the real outcomes: purchases, new donors, qualified leads, phone calls, sign-ups.

That’s why incrementality is the honest form of TV attribution. Incrementality measures whether a conversion would have happened without the ad running. The question isn’t “did more people search for us.” It’s “would this sale have happened if the ad hadn’t aired.” That’s the version that’s hardest to fake, and it’s the one worth building around.

The better question

Most companies are asking, “How do we protect our search traffic?”

A better question is, “How do we get people to ask for us by name?”

The winners in the AI era won’t only be the brands that show up when someone asks an AI a generic question. They’ll be the brands people ask the AI for by name. That’s the difference between catching demand and creating it.