Fresh Step’s AI Slashes Ad Costs 27% in 2026

Listen to this article · 7 min listen

Fresh Step just put up a huge win for AI in marketing, slashing their cost per acquisition (CPA) by 27%. They did it by letting advanced AI handle audience targeting and creative, which fundamentally changes the math on ad costs. Does this mean traditional media buying is officially on the ropes?

Key Takeaways

  • Fresh Step’s Q3 2025 campaign cut CPA by 27% using AI to find better audiences and optimize creative on the fly.
  • AI-generated ad variants beat the human-made controls with a 19% higher click-through rate (CTR), proving the machine could drive better engagement.
  • Predictive analytics let them shift 35% of their ad budget away from failing channels to better ones in real-time, stopping waste before it happened.
  • This worked because their data science team worked directly with marketing. It wasn’t just some off-the-shelf software fix.
  • If you want these results, you have to build your own internal AI know-how and get your data clean. AI isn’t a magic wand.

27% Reduction in Cost Per Acquisition: The Power of Predictive Targeting

That 27% drop in CPA is the number everyone’s talking about, and it came from the AI’s incredibly granular targeting. The system chewed on historical purchase data, site behavior, and third-party demographics to build its own audience segments. A human marketer might target “cat owners, 25-45, suburban.” The AI went deeper, finding niches like “first-time cat owners in urban environments who frequently purchase eco-friendly household products and engage with pet content on specific niche forums.” This precision meant platforms like Google Ads and Meta Ads Manager were serving ads to people almost guaranteed to be interested.

I’ve seen this play out on my own campaigns. This kind of CPA drop comes from finding more relevant clicks. When your targeting is this sharp, you stop wasting money showing ads to people who will never buy. That directly lowers your cost for every single purchase or sign-up. You’re shifting the entire model from broadcasting to a massive audience to having a quiet, effective conversation with the right person.

19% Higher Click-Through Rates on AI-Generated Creatives

On top of the targeting, Fresh Step used AI for dynamic creative optimization (DCO) and got a 19% lift in click-through rates (CTR) over their human-designed ads. This is the part that makes some marketers nervous, feeling like creative is a human-only skill. But the data shows that AI is an amazing tool for testing and iterating through thousands of creative combinations at a speed no human team can match. The machine figured out which headlines, images, and calls-to-action worked for each of those micro-segments. Maybe one group liked funny kitten pictures, while another responded to a clean design that talked up the odor control tech.

This is AI acting as an incredibly fast and tireless creative assistant. Human strategists still provide the initial concepts and assets, but the AI refines, adapts, and deploys endless variations. The lesson is simple: you have to give the machine a lot to work with. A small library of five images will limit its power, but feed it hundreds of photos and dozens of headlines, and the performance speaks for itself.

35% Real-Time Budget Reallocation Based on Predictive Performance

Here’s a benefit people don’t talk about enough: real-time budget reallocation. Fresh Step’s AI dynamically moved 35% of its ad spend away from placements that were starting to fade and into ones with higher predicted returns. This is way beyond the typical weekly or monthly budget meeting. The AI is constantly watching, predicting which campaigns will perform and which are about to hit a wall. If an ad set on one platform starts showing diminishing returns, the AI automatically pulls budget and sends it somewhere else that’s crushing it. This happens in hours, sometimes minutes.

So many companies are too slow here. Manual budget shifts are reactive, usually happening long after you’ve wasted a ton of money on something that wasn’t working. AI’s predictive ability lets you proactively channel money where the returns are highest, maximizing every dollar. Getting this to work requires connecting all your ad platforms into a single data stream and having a solid attribution model so the AI knows what’s actually driving results.

Beyond the Hype: The Important Role of Data Hygiene

The numbers from Fresh Step are great, but it’s easy to get the wrong idea about how they did it. You can’t just buy an AI tool and expect these results. The real secret, the part nobody wants to talk about, was Fresh Step’s intense focus on data hygiene and integration. Before a single algorithm was turned on, they spent six months cleaning, structuring, and connecting their own first-party customer data with third-party sources. This meant standardizing customer profiles, killing duplicate entries, and setting up strict data governance rules.

If you feed an AI messy, unreliable data, you will get useless results. Fresh Step knew their AI’s predictions would only be as good as the data it was trained on. This upfront investment in data infrastructure is the hardest, most tedious part of any AI project, but it’s completely non-negotiable if you want the kind of efficiencies Fresh Step saw. Any company trying to copy this playbook needs to get serious about data readiness first.

Fresh Step’s playbook shows that AI isn’t a magic bullet. It’s a powerful tool that works best when built on clean data and guided by smart people. The main takeaway is to get your data infrastructure and internal AI skills in order now, because your competitors are already doing it and seeing massive savings. For any CMO wanting to do the same, figuring out how to boost ROI with data is the first step. And don’t forget that AI data ethics and compliance are not optional, they protect you from huge fines and keep your customers’ trust.

What specific AI technologies did Fresh Step use for their ad campaigns?

They used a mix of machine learning for predictive audience building, natural language processing (NLP) to analyze ad copy, and generative AI inside a dynamic creative optimization (DCO) system to assemble the final ads.

How long did it take Fresh Step to implement their AI advertising strategy?

The setup took about six months, most of which was spent on data cleaning and integration. After that, the campaign ran for one quarter to collect the performance data.

Can small businesses achieve similar AI-driven ad savings?

Yes, the same principles work. Smaller businesses can get started with the AI features already built into platforms like Google Ads Smart Bidding or Meta’s Advantage+ campaigns. The key is to focus on the quality of your own customer data.

What were the biggest challenges Fresh Step faced during AI implementation?

Their main hurdles were stitching together different data sources, making sure everything was compliant with privacy laws, and training their marketing team to work alongside data scientists and actually use the AI’s recommendations.

Is human oversight still necessary when using AI for advertising?

Absolutely. A human still needs to set the strategy, provide the creative vision, and make the final call, especially on ethical questions. The AI is there to automate and optimize the execution of that strategy at a massive scale.

Editorial Team

The Editorial Team at CMO News Time is a group of writers and editors who cover marketing, brand, and growth. They research each story and publish under a shared desk byline — experts in their fields, working as one team.