Brand Tracking Guides The Split Between CTV Reach And AI Creative As Production Costs Fall

AI tools now make new ads fast and affordable to produce, giving brands a choice between putting more money toward reach or toward creative. That choice should depend on what's keeping consumers from buying. If consumers don't know a brand exists, reaching more of them through channels like CTV is the better investment. If they know the brand but don't see it as a fit, better messaging will move them further than more exposure.

Laura Kavanagh is SVP, Head of Media at Mediaplus North America, the U.S. arm of Mediaplus, an independent media agency network that's part of Germany's Serviceplan Group. She has spent about 25 years in media planning, including leadership roles at Canvas Worldwide, Mediahub, MediaCom and Starcom, working with brands such as P&G, Subway, and AT&T. Since joining Mediaplus in 2022, she has helped build the U.S. agency, where her team works with brands in food, tech and B2B to identify what's holding back their sales and plan media spending around it.

"If you have high reach and the creative is weak, you're probably going to alienate people. If you have low reach and a ton of creative, no one's going to see it. It's finding that balance," Kavanagh says. Brands have to weigh which message consumers should see against how many of them need to see it. For a newer brand, reach usually comes first, while an established brand may get more out of sharper messaging.

Awareness or relevance

Marketers can see what's limiting sales through brand tracking studies, which survey consumers over time about whether they know a brand and how they feel about it. The results shift as a brand matures, so the split between reach and creative needs another look whenever the data changes. "If people are aware but don't find you relevant, you might want to focus more on the creative messaging. If your brand health metrics are telling you that people are unaware, then maybe you need to spend a little more time on reach," she says.

For brands with a relevance problem, marketers can test many ad versions to find a message that works. AI tools and dynamic creative, which changes images, copy or offers for each viewer, let teams make far more versions than a single shoot can. Each consumer can then see the version most likely to connect, and the results show which messages lead to sales. "The ability to more efficiently produce assets and have a variety of assets at a much quicker pace is key, and it helps ensure the media we're putting in front of people is relevant," Kavanagh notes.

More ad versions also help on social platforms, where algorithms now handle much of the targeting. On Meta, marketers increasingly reach broad audiences and let the platform find likely buyers. The algorithm can read what's in an ad, such as a teenager drinking a soda, and show it to consumers likely to respond to that scene. That means the ads a team creates help decide who sees them, so creative choices shape reach too. "The algorithms have gotten so much better that social has become less about media targeting and more about the creative strategy," she adds.

Test, then spend

Marketers with dozens of versions of an ad still need to know which ones are worth paying to run. Pretesting tools show these ads to AI-generated personas or to human panelists and score how each is likely to perform. Results come back faster than a focus group and can cover a much larger group of people, so teams can check a big batch at once. "We have a couple of tools where we're able to pretest creative before it goes to market really quickly without having to spend a lot of money," Kavanagh says.

Teams can run these tests as early as the storyboard stage, so weak ideas get dropped before anyone pays to film them. Marketers also learn which messages work before buying media, so more of their spending can go toward putting proven ads in front of consumers.

After the ads launch, proving the spending worked gets harder when a campaign runs across CTV, social, search and digital video. Large platforms such as Meta and Google, often called walled gardens, keep their data inside their own systems, so their results can't be merged into one dashboard to see where audiences overlap.

Counting each sale once

Platform reports can make a campaign look healthier than its sales suggest. When a brand runs ads on both platforms, each may count some of the same sales, so the campaign can hit its return-on-ad-spend goal while sales stay flat. Incrementality tests, which pause or change ads in some markets and compare sales with markets where ads keep running, show which spending is driving results. Marketing mix modeling adds a longer view of how much each channel contributes. "There's no way to understand the actual impact without doing that level of diagnosis," Kavanagh says.

AI also gives planners more time for that work. Data analysis that once took up much of their day now moves faster, leaving room for questions about who the consumer is and what the business needs to grow. "There's less legwork, and that gives us the ability to think more strategically about the business as a whole, and to think of bigger media and more creative ideas that help the client stand out," she says.