Pacsun has built, by hand and out of teenagers, the instrument that AI vendors spend years selling. Its co-creation engine, a standing youth council and a 6,000-person Youth Report, reads what its customers will buy before they buy it. That is the forward signal trend-forecasting houses sell by subscription. The crowd is cheaper, and Pacsun’s results argue it is also more accurate.

The mechanism is deliberate. Last September the brand seated a council of young creators at what it calls the leadership table, with direct access to the decisions its executives make. A focus group is convened to admire the line; this council is asked whether it will sell. The answer is a forecast, drawn from the people it is built to sell to.

In January the loop got an interface. The PS Community Hub is a shoppable app where users post, sell, earn commission, and, if they perform, are invited to design the next capsule. Every tap and purchase is logged as it happens, not surveyed once a season. The app runs on AI for moderation and personalisation. The forecasting is the one job Pacsun kept for people.

Set that beside what the vendors sell. Heuritech reads millions of social images, tags 2,000 attributes, and projects a trend up to 24 months out. It is an elaborate way to infer a want from the pictures it leaves behind. The vendor infers desire from strangers; Pacsun’s council never has to be inferred, because it volunteers.

Pacsun prefers a warmer word for the instrument. A senior merchant speaks of collective thinking and inspiring youth to create a better future, the vocabulary of a movement, not a forecast. The vocabulary earns its keep. A demand model is easier to fund when it reads as a community.

A trend forecast is what a brand buys when it has lost the nerve to ask; Pacsun asked.

The arithmetic favours Pacsun. Sales reached $970 million in 2025, per SGB Online citing a Needham & Co. research note, on 10 straight quarters of positive comps, eight of them double-digit, with EBITDA margin up three points since 2023. That is not a marketing budget paying off — it is a forecasting budget that was never spent. The council costs mentorship and travel. The app leans on engineering the company already owns.

Harvard has made the engine a case study, and Pacsun’s chief executive has a book out about it. The classroom will read it as a lesson in culture. The colder lesson is commercial: a brand can read its own demand by paying the people who are the demand. That leaves the trend vendors selling, by subscription, a copy of the answer the customer hands over for a commission.