AI-first.

By 2024, running brand at PRISM meant coordinating multiple in-house teams, dozens of agency partners, and creative output across thirty-five countries.

The old model — one built when the company was one brand in one market — was quietly becoming the thing that slowed everything down.

I rebuilt the entire operating model around AI-first workflows. Not "we use ChatGPT now." A structural rebuild: what work stays external, what comes in-house, how briefs move, how creative gets made, who reviews what, and how a piece of work goes from insight to shipped asset without needing four sequential meetings.

Any brand leader can hit an eighty percent cost-reduction number if they're willing to torch the team on the way there. The interesting version of this work is doing it while the people on your team choose to stay — because if you can't, you haven't rebuilt an organisation, you've just decimated one.

Ninety percent retention over two years means the people on the team saw the shift as an upgrade to how they worked, not a threat to whether they had a job. That reframing was the actual leadership challenge. Everything else — the tools, the workflows, the vendor contracts — was execution.

The core moves:

• Insourcing what agencies were charging most for. Not everything — the highest-craft work still lives outside — but the middle sixty percent of the brand output that agencies were charging premium day rates for, without adding proportional value, came in.
• AI-native brief-to-asset workflows. Rewriting how a brief becomes a first draft, how iteration happens, how review cycles work. AI does not replace the strategist or the creative lead. It removes the six-hour first-draft wait between them.
• Modular creative systems. Building brand assets as composable modules that regional teams can localise without breaking the brand. This is what makes thirty-five countries operable from a lean core team.
• Vendor rationalisation. Fewer partners, deeper relationships, retained work rather than per-project.
• Redefining what the in-house team actually does. Less operator, more editor. Less production, more direction.

The concrete outcome is that PRISM's brand organisation can now ship more work, in more markets, at higher quality, at twenty percent of the historical cost. That is a structural capability change, not an incremental efficiency gain.

The strategic outcome is that brand is no longer the bottleneck for the company's expansion. New brand launches — like Sunday — new market entries, new campaigns, new sub-brands — all can move at the pace of the operator strategy, not the pace of the creative supply chain.

If you're doing this work: start with the retention question, not the cost question. Design the new organisation around the people you want to keep. The cost gains follow.