Fresh Step’s 90% AI Ad Cost Cut in 2026

Listen to this article · 10 min listen

The pet care market is a bloodbath for digital marketers trying to get customers efficiently. Fresh Step, the cat litter brand, just pulled off something pretty remarkable in this space, using AI ad creation to cut their advertising costs by a staggering 90%. Let’s break down exactly how they did it, because this shows how automation can seriously improve campaign performance and actually deliver on the promise of cost-effective campaigns.

Key Takeaways

  • Fresh Step’s ad costs dropped 90% after they brought in AI for creative generation and dynamic targeting.
  • The campaign spit out over 1.2 million unique ad variations, which boosted click-through rates 3.5x compared to their old manual efforts.
  • By targeting tiny audience segments with personalized messages, their cost per conversion for new customers fell by 60%.
  • The AI’s feedback loop was constantly running, optimizing creative and bid strategies almost hourly based on live performance data.
  • If you’re going to use AI tools, you have to get your data integration and campaign objectives sorted out first to have any hope of improving ROI.

The Challenge: Overcoming Ad Saturation in Pet Care

Before this AI project, Fresh Step was hitting the same walls as everyone else in the pet supplies game. Ad costs kept creeping up, and you could tell audiences were sick of generic ads. Their old campaigns were fine, but they were stuck in the slow, expensive cycle of traditional A/B testing. The goal was simple: get the cost per acquisition (CPA) for new cat owner subscriptions down, without losing (and hopefully gaining) conversion volume. Their Q3 2025 numbers, with an average CPA of $45.20, was a figure that management knew was a dead end for growth.

Initial Campaign Metrics (Q3 2025 Baseline)

  • Budget: $250,000 per month
  • Duration: 3 months (July 1 to September 30, 2025)
  • Average CPL (Lead): $18.50
  • Average ROAS (Return on Ad Spend): 1.8:1
  • Overall CTR (Click-Through Rate): 1.1%
  • Total Impressions: 150 million
  • Total Conversions (New Subscriptions): 5,531
  • Average Cost Per Conversion: $45.20

Strategy Shift: Embracing Generative AI for Creative at Scale

Fresh Step’s big strategic pivot in Q4 2025 was to go all-in with generative AI. They didn’t just dip their toes in. They partnered with an ad tech platform that specializes in AI-driven creative, aiming to do more than just basic dynamic creative. The system wasn’t just swapping out headlines. It was generating completely new visuals, copy, and calls-to-action (CTAs) on the spot, all tailored for hyper-specific audience segments.

The AI-Powered Creative Process

The marketing team basically dumped their entire library of assets into the machine, product images, lifestyle photos of cats, customer reviews, even competitor ads for context. The AI then started churning out thousands of ad variations. It would mix and match visual elements, copy angles, and emotional tones. So, an ad for someone the system pegged as a “first-time cat owner, concerned about odor” would get a close-up of a happy cat with a headline about “advanced odor control.” Meanwhile, an ad for a “multi-cat household, budget-conscious” shopper would see a totally different visual emphasizing “long-lasting value.” You just can’t get that granular at scale with a human team.

The system also digested all the historical data from Fresh Step’s Google Ads Performance Max campaigns and Meta’s Advantage+ Shopping Campaigns. This let it predict which creative elements, like color schemes, what’s in the image, or specific words, would hit home with different groups of people. It was constantly iterating, learning from every click and conversion to sharpen its next batch of ads. This feedback loop, running in hours instead of weeks, is where the real magic happens.

Targeting Precision: Micro-Segments and Behavioral Triggers

The AI didn’t just build the ads, it also completely changed Fresh Step’s targeting. They moved away from broad interest groups and let the platform build micro-segments from a huge mix of data points, including purchase history, browsing patterns, and even sentiment from social media chatter about cat care. For example, the system could identify someone who recently searched for “kitten supplies” or read articles about cat health. Could a human marketer manage that many tiny segments? No way.

Another big piece of the puzzle was using predictive bidding strategies. The AI would dynamically shift bids for each little segment based on its real-time probability of converting, which meant it was constantly optimizing how the budget was spent across platforms like the Google Display Network and Instagram. The money always flowed to the most promising impressions, instead of being spread evenly or based on old averages that are useless by Tuesday.

Campaign Execution: A Look at the Data

They ran the AI-driven campaign for three months, from October 1 to December 31, 2025. The results showed up almost immediately. Over the course of the campaign, the system generated more than 1.2 million unique ad variations, a scale that’s just not possible with traditional workflows. This constant refresh is what helped them blow past ad fatigue and keep engagement high.

Q4 2025 AI-Driven Campaign Metrics

  • Budget: $250,000 per month (consistent with baseline)
  • Duration: 3 months (October 1 to December 31, 2025)
  • Average CPL (Lead): $3.70 (an 80% reduction)
  • Average ROAS: 9.5:1 (a 427% increase)
  • Overall CTR: 3.8% (a 245% increase)
  • Total Impressions: 180 million
  • Total Conversions (New Subscriptions): 54,054
  • Average Cost Per Conversion: $4.62 (a 90% reduction)

That 90% cost reduction in ads wasn’t just a vanity metric. For the same budget, it produced a tidal wave of new subscribers. When your cost per conversion drops from $45.20 to just $4.62, you can scale up your acquisition efforts in a massive way without having to go ask the CFO for more money.

What Worked Exceptionally Well

  1. Hyper-Personalized Messaging: Matching the right creative to specific audience needs was the main reason CTR and conversion rates shot up. The ads felt more like solutions to problems than generic marketing noise.
  2. Rapid Iteration and Optimization: The AI’s learning loop meant bad ads were killed off quickly and replaced with better versions. This agility is what kept the campaign running hot.
  3. Dynamic Bid Management: Predictive bidding made sure the budget went to the highest-potential impressions, squeezing maximum value out of every dollar.
  4. Cross-Platform Teamwork: The AI managed creative and bidding across different ad networks, creating a cohesive experience for the user no matter where they saw the ad.

Challenges and What Didn’t Work (Initially)

It wasn’t all a cakewalk. Early on, some of the AI-generated ad copy was weirdly off-brand, requiring a human to step in and tighten the AI’s guardrails. Some headlines were too pushy or used slang that just didn’t fit Fresh Step’s voice. This just shows that you need human oversight, especially when you’re first “training” the AI. In my own experience, this is the step marketers always underestimate, the time it takes to teach an AI the subtleties of your brand and what not to say.

Data integration was another early headache. Fresh Step’s data engineers had to put in serious work to get all the first-party and third-party data sources to play nice with the AI platform. A successful AI marketing project is built on a clean, unified data feed. Without it, the AI’s predictions would have been garbage. It’s the unglamorous, foundational work that everyone forgets but that makes or breaks the whole thing.

Optimization Steps Taken

To fix the brand tone issues, the team set up much stricter “brand safety” rules inside the AI platform. They fed it more explicit examples of approved and disapproved copy, and the AI learned to stay within those lines. They also had a content manager spot-check a sample of AI ads every day, giving direct feedback to the system.

On the data side, they built a proper data pipeline with cloud-based ETL tools. This ensured their CRM, website analytics, and the AI ad platform were all synced up in real time. Cutting down that data lag meant the AI could react to what a user was doing almost instantly.

The Future of Ad Creation: A Human-AI Partnership

Fresh Step’s success really shows the potential of AI in advertising. It proves that AI ad creation offers a level of personalization and efficiency that was impossible just a few years ago. A 90% cost reduction isn’t a one-off miracle. It’s what’s possible in this new era of marketing. This case study also shows how the job of a marketer is changing. We’re not just manually building every ad anymore. We’re becoming the strategists, data analysts, and AI trainers who guide the technology to get these kinds of results. The future is a collaboration where our human creativity and intuition direct the raw power of the machine.

How does AI reduce ad creation costs so drastically?

It automates the generation of countless ad variations, runs A/B tests at a massive scale, and optimizes targeting and bidding in real time. This gets rid of most of the manual production work and funnels the budget to high-performing ads and audiences, which directly lowers the cost per conversion.

What types of data are important for successful AI ad campaigns?

You need a rich mix of data: historical campaign performance, customer demographics, behavioral data like website visits and purchase history, a full library of creative assets, and even intel on your competitors. The cleaner and more complete the data you feed the AI, the better it can learn and optimize.

Is human oversight still necessary when using AI for ad creation?

Absolutely. People are still needed to set the overall strategy, define brand guidelines, act as a brand safety backstop, interpret complex results, and give qualitative feedback to the AI. The AI is a powerful engine, but it needs a human driver to tell it where to go and keep it on the road.

How can smaller businesses implement AI for cost-effective ad campaigns?

You can start by using the AI features that are already built into platforms like Google Ads (Performance Max) and Meta Ads (Advantage+). These incorporate AI for bidding and creative suggestions. There are also more accessible third-party AI creative tools with tiered pricing that can help automate basic ad creation.

What are the potential pitfalls of relying too heavily on AI for advertising?

If you’re not careful, you can lose your brand’s voice, create ethical problems with overly intrusive targeting, or get stuck in a rut where the AI only optimizes for what’s worked before instead of finding new breakthroughs. There’s also the “black box” problem where you don’t know why the AI is making certain choices. You have to balance the automation with human strategy.

Editorial Team

The Editorial Team at Creative Ads Lab 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.