Generative AI is reshaping how legacy consumer brands create, test, and launch creative assets. This is especially true when these ads need to be produced at scale, and quickly, for social platforms.
The Wall Street Journal recently profiled Clorox’s five-year, $580 million digital transformation, looking at how teams from Hidden Valley Ranch to Burt’s Bees have woven AI into everyday workflows. Their experience offers a real-world stress-test of enterprise AI.
Spoiler: Clorox turned to Microsoft Copilot for most of their GenAI needs. We’re a little biased, but purpose-built stacks such as The Marketing Cloud can achieve repeatable wins far faster than off-the-shelf tools like Copilot, ChatGPT, or any instances of dreaded BYOAI.
Below, we’ve condensed 3 key learnings from Clorox’s AI journey.
AI doesn’t hit 100% of ideas out of the park, but then again, neither do your human colleagues.
As the WSJ reports, one Copilot render for a Hidden Valley Ranch spot—featuring decidedly limp and pasty chicken wings—was an unappetizing dud. Instead of shelving the tech, the team tightened prompts, injected brand-safe assets, and built a quick red-yellow-green scoring model so anyone could specify if outputs were 'post-worthy' or 'try again.'
The AI tools were soon spitting out on-brand variations in the time a designer once spent on a single mock-up, cutting creative production costs while increasing test velocity.
The lesson: Treat generative AI like an eager junior designer—give it structure, critique ruthlessly, and watch quality incrementally improve.
It’s important to note, though, that working with an off-the-shelf product like Copilot can require a great deal of prompt-engineering expertise and experimentation in order to reach your marketing sweet spot. Trialing a purpose-built tool like Propellers might get you to the winner’s circle a little earlier.
Clorox leadership challenged every function to find its own AI win. R&D brainstormed a fizzing “Toilet Bomb” cleaner, while customer-insight teams used AI analysis to mine tons of reviews, learning that scent—not color—drives Burt’s Bees lip-balm love.
The lesson: A grassroots approach bred rapid adoption and a steady stream of small victories that, together, delivered big impact—without a single AI-related layoff.
Empowerment beats edicts, especially when the tech is this new. It’s the way to get buy-in from your teams, rather than having them feel like they’re being backed into a corner and forced to adopt new tech they don’t yet believe in. So yes, promote and advocate for AI adoption, but let your teams experiment with the tech to stoke enthusiasm organically.
Clorox doesn’t chase AI novelty; it chases ROI. Savvy brands track concrete KPIs—cost-per-creative, time-to-insight, incremental revenue from new product ideas—and kill experiments that don’t move the needle.
The lesson: Grounding experimentation in numbers keeps budgets safe, hype in check, and ever-important C-suite support locked in. The Marketing Cloud aims to do the same—surfacing campaign lift, channel efficiency, and predictive next steps—so your AI program stays tied to results.
While measuring the ROI of new AI tools can be a slippery business, it’s essential if you want to concrete proof that this tech is improving workflows. Vibes alone won’t cut it. Tools like BERA.ai’s Brand Explorer can also help quantify the financial impact of AI-driven brand investments, forging a closer relationship between your CMO and your CFO.
Clorox’s AI adventure proves three things: speed and iteration beat perfection, empowerment beats mandates, and ROI beats novelty.
Brands that internalize those principles—and deploy them inside a purpose-built platform—are setting themselves up for AI adoption that's actually making a difference, not just chasing trends and buzzwords.
The Marketing Cloud is built for exactly that mission, whether you’re looking to optimize creative assets pre-flight, or leverage AI to run marketing campaigns in real time.
Want to turn AI curiosity into a competitive edge? Register for free and explore the platform today, and request a demo when you’re ready.
