An American cat litter brand produced three TV spots in 30 days, at roughly one-tenth the usual production cost. All with AI.
The story, recently covered by Marketing Dive, isn't really about technology. It's about what you do with the nine campaigns you can now afford for the price of one.
Fresh Step, a brand in Clorox's portfolio operating in a $5 billion U.S. cat litter market, worked with AI-powered agency RECE to create a full campaign featuring AI-generated talking cats. Fifteen-second spots, from brief to airdate in under a month. At a cost that would normally cover pre-production alone. As a strategist, the number that catches my attention isn't the 90% saving. It's what happens to the other 90% of the budget.
Cheaper production is not a strategy
The first instinct when production costs drop 90% is to spend less. The second, slightly smarter instinct is to produce more for the same budget. Neither is actually a strategy.
What makes the Fresh Step campaign interesting isn't the price tag. It's the architecture. Eric Schwartz, Clorox's CMO, describes the approach as a system: each spot can be adjusted, each message tested, each variant optimized based on in-market data. The brand didn't bet on one creative idea. It bet on a testing framework. That's a fundamentally different strategic choice.
The difference shows up in the decision cycle too. In the traditional model, a CMO approves a concept, allocates the production budget, and waits 3-6 months to see if it works. With AI production at a fraction of the cost, you can launch 3 variants in the first month, read the performance data, and invest more in whichever one performs. It's no longer a bet on a single idea. It's a controlled experiment with real data in real time.
Most campaigns still operate on the old model: one big idea, one big budget, one big hope. If it works, great. If not, change the agency. A modular approach changes that dynamic entirely. You stop betting everything on a single execution. You test, adjust, iterate. But that requires a strategy built for iteration, not a brief that asks for "something viral".
The real question: what do you do with the freed-up budget?
When production cost $100,000 per spot, every creative decision was high-stakes. It had to work the first time. There was no room for experiments.
But if the same production costs $10,000? Suddenly you have budget for 10 variants. Or for 3 variants tested across 4 different audience segments. Or for continuous content instead of campaign bursts. When AI is already generating a billion dollars in ad revenue, the question is no longer whether the technology works. It's how to integrate it intelligently.
Fresh Step chose exactly this path: what Clorox calls "systems of relevance." An approach where content is produced and distributed continuously, not in campaign waves. Every asset is treated as a hypothesis, not a finished product.
The implication for brands is direct. It's no longer about "let's run a Black Friday campaign." It's about "we have a communication system that works year-round, and Black Friday is just a moment where we amplify the messages that already perform." But that requires a mindset shift at management level. Moving from "approve a creative concept" to "approve a testing framework" is harder to sell in the boardroom. But the numbers help: Deb Crandall from Clorox mentions that the savings from a single campaign could fund roughly 10 more productions. Instead of one creative execution per quarter, you have material for continuous testing throughout the year. The constraint was never a lack of ideas. It was the cost of testing them.
Transparency as a strategic decision
A detail that's easy to overlook: Fresh Step doesn't hide that its ads are AI-made. Deb Crandall, senior marketing director at Clorox, says explicitly that the cats in the spots are clearly AI-generated and the brand isn't trying to fool anyone.
That's a choice. And it's a smart one.
In 2026, consumers aren't naive. On Meta, AI already makes the ads, and platforms don't even hide it. If you try to pass off AI production as traditional, the backlash risk far outweighs any perception gain.
Fresh Step turned a potential vulnerability into a brand element: animated cats with different accents and attitudes. It's funny, it's memorable, and it doesn't pretend to be something it isn't. For brands considering AI in creative production, the lesson is straightforward: transparency isn't a concession. It's positioning. And in a market where every competitor is weighing the same AI adoption decision, being first to own it openly is a real competitive advantage.
The brand-agency relationship is being rewritten
If creative production costs 10 times less, the agency's role changes fundamentally. You're no longer selling production hours. You're selling strategy, frameworks, testing capacity, and data interpretation.
Agency RECE did exactly that: it proposed a system, not a set of spots. And Clorox bought that system because it had a brand strategy that supported it. In practice, this means agencies still monetizing through production markup have a problem. Your ad account might look great, but if the strategy behind it isn't solid, cheap production just accelerates the distribution of mediocre messages.
For agencies, the implication is direct. If your value as an agency sits in execution, AI just compressed that value to a tenth. If your value sits in strategy, in the ability to build and interpret a testing system, then AI just gave you the best possible sales argument: you can demonstrate results faster, with less risk for the client.
The right conversation between brand and agency is no longer "how much does a spot cost?" It's "how many variants are we testing this month and what are we learning from them?"
AI didn't just reduce creative production costs. It removed the main excuse brands had for not testing more: the budget. Now that production is no longer a barrier, the only real question is whether your strategy is good enough to know what to test. And, more importantly, what to do with what you learn.
