One launch. 100,000+ personalized films.

The business problem

A global category leader was launching a new product into a market of hundreds of thousands of customers spread across regions, languages and buying seasons. Mass media could build awareness, but awareness was not the constraint - relevance was. The launch had to make each customer feel the product was for them specifically, and no format does that better than film. The problem: at six figures of audience, personal film had always been economically impossible.

The audience challenge

The audience did not share a language, a region or a use case. A single national film would speak to an average customer who does not exist - polite, general and forgettable. Worse, this audience makes considered, high-trust purchase decisions on advice from people who know their name. The bar was not attention. It was the feeling of being addressed.

The strategy

Make the film one to one, and treat it as a delivery system rather than an ad. One idea worth personalizing, then a system that could carry it to each individual: the customer's name spoken aloud, their language, their region's context, the product that mattered to them - delivered to the device already in their hand. The strategic bet was simple: relevance is the price of attention, and a personalized film pays it up front.

The production system

This is where the work was won. A master film was built for variation from the first storyboard - fixed parts that never change, variable parts designed to be generated. Clean first-party data was mapped to those variables, with real time spent removing the gaps that would otherwise produce a film addressed to "Dear Valued Customer". An AI production pipeline then generated voice and on-screen text natively per language, assembling each version file by file, with human quality gates so the hundred-thousandth film held the standard of the first. The full pipeline is described in how we make 100,000 personalized videos.

The distribution model

Personal film deserves a personal channel. Each film was delivered straight to the customer's phone - WhatsApp-first, in the thread where a name and a known sender mean something - rather than broadcast into a feed. Delivery was sequenced by region and season so the film arrived when the decision was live, not weeks before or after.

The output

100,000+ unique films from one launch. Each one addressed to a real person by name, in their own language, about the product relevant to them - produced through one pipeline, to one standard, at a fraction of what traditional production economics would demand for even a hundredth of that volume.

The outcome

A step-change in message recall and response against the brand's previous mass campaign, measured directionally across the launch window. As importantly, the launch left the brand with an owned capability, not a spent budget: the master-film-plus-pipeline system can be rerun for the next product, the next season, the next market.

What made it scalable

Three decisions. The master film was designed for variation before a frame was produced, so scale never fought the idea. The data was cleaned before it touched production, so volume never amplified errors. And personalization was built at the production layer, not bolted on at the targeting layer - the creative itself changed per person, which is the difference between a mail merge and a message. This is what we mean by film production built for systems.