March Madness Mascots

One test for ESPN. Then we kept going.

ESPN wanted to see what AI-native production could do with a mascot. We made them Smokey.

March Madness was already underway. The tools were in hand. One mascot wasn't enough. So we kept making them. Live, through the tournament. Every matchup, every upset, every cinderella run got its own aged-up portrait or motion piece. Eight final portraits. Thirteen motion pieces. Every frame produced within hours of the real game happening.

Mascots aged like the athletes they represent.

Every March Madness produces the same creative layer: bracket graphics, team-pride edits, upset reels. Competent, forgettable, always late. Production lead time is the bottleneck. By the time a brand idea clears approval, the tournament has already decided its own story.

Invert the constraint. Aging-athlete tropes, applied to mascots. Smokey at the end of a long career. The Gator, drawl intact, holding court. The Duke Blue Devil, dignified in decline.

The method was the message. Traditional live-event work plans for every possible outcome and lets the bracket choose among them. AI-native production lets the bracket write itself, then answers each game with its own piece. The creative stays relevant because it is made inside the event, not before it.

The constraint was the brief.

Every brand team watches March Madness go by and wishes they'd done something. By the time they get approval to produce, the bracket is already filled in. The campaign has already missed its window.

This project is a demonstration of what changes when production no longer needs lead time. If a piece can ship in four hours, it can ship in the right hour. The creative question becomes "what's the idea" not "can we get it approved in time."

Traditional production would cost six figures for a series like this. We did it with a laptop and a tournament on a second screen. Not because cheaper was the goal. Because the cadence of the idea required it.