Animating Marriott's Logo Into a Dozen Christmas Worlds With AI, Not Keyframes

Not every Marriott project we have worked on is the Four Points Flex configurator. This one was seasonal, smaller in scope, and solved a completely different problem: placing Marriott's "M" logo inside a range of festive environments for a global holiday campaign, without hand-animating every single scene.

A traditional approach to this brief would mean picking two or three festive settings, hand-modelling or filming each one, and animating the logo into each individually. That gets expensive fast, and it caps the campaign at however many scenes the budget covers. Marriott operates in dozens of markets with very different ideas of what "festive" looks like: snow reads as Christmas in some regions and as nothing at all in others. A campaign built around two or three hand-built scenes cannot flex to that.

We used AI generation to produce a much wider range of festive environments for the logo to sit inside, then handled the placement and motion work to make the "M" feel genuinely part of each scene rather than pasted on top of it. That second part is where the actual craft sits. A generated background and a clean logo animation do not automatically belong together: lighting has to match, shadows have to fall the right way, and the logo's own motion has to respect the perspective of a scene it did not originate in. Getting an AI-generated environment and a brand mark to sit convincingly in the same shot is closer to compositing work than to prompting.

The payoff was range. Instead of a campaign locked to two or three markets' idea of Christmas, Marriott got variations that could flex by region and by placement, snow-heavy scenes for some markets, warmer festive settings for others, without the cost of filming or fully modelling each one from scratch.

It is a narrower use of AI than the pre-viz work we have written about before, where AI helps sell a project before it exists. Here the AI's job was pure production efficiency: generate more raw material than a traditional shoot or build could afford, so the animation and compositing team could spend their time on the part that actually needs a human eye, making it look like it was always one shot. That division of labour, AI for range, people for the compositing that convinces, is quickly becoming the more interesting story in motion work than generation on its own.