On average a CMO gets around five to 20 martech and AI solutions pitched every day, depending on company size. And at enterprise scale, the level Unilever finds itself at for example, this is often more, running into the 30s. Add in existing partner proposals and the “self‑discovery” every curious marketer does on their own, and the signal‑to‑noise ratio gets really really overwhelming.
This too is a problem for Unilever’s associate director of digital capabilities, Rachelle Schuimer, who spends her weeks swimming in vendor pitches. So how does she separate a shiny capability from a real solution? And how does any marketer scale content without sacrificing creative quality or overwhelming teams?
Start with the right lens
For Rachelle, who recently spoke to Celtra’s Vanja Brzin about this problem in a live session titled Behind Unilever’s AI-powered creative transformation, the first move is distinguishing “what’s available” from “what we need.” That means asking four questions up front:
- Strategic fit: Does this address a genuine pain point?
- Scalability: Can it work across brands, categories, and markets?
- Impact: Will it make marketers’ lives meaningfully easier?
- Time horizon: Does it solve a short‑term churn issue or align to a longer‑term bet on where marketing is going?
This framing, Rachelle explains, reduces impulse buys and re‑centres decisions on outcomes, not features.
Plan to fail and learn
Unilever runs new capabilities through a deliberate test‑and‑learn phase with partners. Some won’t perform as hoped. That’s not a drama at one of the largest FMCG firms in the world, it’s the cost of learning. Their mantra is setting clear success criteria upfront and being honest, early, and often with partners about reality vs. expectation.
A lot of what Rachelle does is making a call early on whether to pivot based on what the team learned or pull the plug and redeploy time, energy, and budget. “Not everything will be a success,” says Rachelle. But “if you never fail a test, have you tested enough?”
Consider becoming a paid-for member of our community and support our mission. Get extra content, early event invites and more.
Unilever’s US team produces around 22,000 assets in a given year. Though that is a high number, the point isn’t the figure, it’s that it signifies the shift in consumer behaviour and media consumption. To stay relevant, brands need more creative variations, fresher messaging, and better audience fit. In that context, creative automation isn’t a vanity play, it’s a necessity.
But Rachelle is clear when she says that production is “the base of the pyramid”, not the endgame. Quantity without quality is noise. Quality without quantity risks generic creatives that under-serve sub‑segments. The flywheel forms when quantity and quality work together. When teams are using volume to learn and then feed those insights back into the creative craft.
The pyramid of needs
Rachelle frames Unilever’s approach as a layered “Maslow pyramid” for creative automation. The basic layers are needed before the higher layers can deliver results. “At the base level, we’re looking at production – can we actually make the number of assets we need?” Rachelle says. “And can we keep creative hygiene intact? Are logos visible, copy readable, nothing awkwardly cropped?”
As Unilever scaled volume, weak points in the workflow surfaced: fragmented steps layered on top of one another, approvals multiplying, brand‑safety checks scattered. The second layer, then, is continuous workflow optimization. This means reducing handoffs, clarifying ownership, and streamlining the path to live. Further up the pyramid is measurement. “Do we have the right KPIs? Are we distilling insights, not just reporting performance?” The goal is to feed those insights back into briefs, templates, and creative decisions to steadily raise quality over time.
Make partnership part of the system
You don’t buy true transformation off the shelf. Rachelle emphasises the need to bring product partners inside the problem. She shares workflows, constraints, and the organisational realities that shape adoptions, such as brand safety, governance, and the many stakeholders involved in enterprise creative.
From there, she says they co‑create a shared vision and roadmap. It’s okay – useful, even – for the brand to challenge the vendor’s product roadmap. And it’s vital that vendors understand the brand’s friction when it comes to real operations, and that the conversation doesn’t just focus on the theoretical potential of the tech.
Define success in three dimensions
Rachelle’s success criteria span craft, commercial impact, and human value. First, creative quality: are assets performing better, and are they firmly on brand? Second, business outcomes: are Unilever’s teams deploying media spend more efficiently and competing more effectively in market? And finally, marketer enablement: have day‑to‑day lives become simpler, with more time for strategy and storytelling rather than focussing low‑value formatting work?
Top‑down and bottom‑up momentum helps here. Rachelle and her colleague’s central guidance sets standards but the teams on the ground surface insights and edge cases which they can’t foresee. Together, with clear communication and standards, they learn, adapt and adopt.
Time saved is creativity earned
Scale isn’t necessarily sexy when it’s invisible, but it is when you see its compounds. If you have 22,000 assets and you shave minutes off the path to “live,” the reclaimed hours across marketers and agencies are enormous. That time can be reinvested. The lesson from Unilever is simple but deceptively hard to execute; treat production as a foundation, not a finish line and design workflows that honour brand safety and speed.
But maybe more so than these, it is about building partnerships that share problems and work in unison. Quality at scale happens when internal systems are designed to learn.
