At first glance, the machine looks almost underwhelming. Inside a brightly lit laboratory at On’s Zurich headquarters, a robotic arm rotates a plastic model of a human foot. At the same time, a nozzle continuously sprays 1.5 km (0.93 miles) of thin molten plastic with almost hypnotic precision.
Layer by layer, the filament wraps itself around the foot mold, gradually forming the upper of a running shoe. The entire process takes about three minutes.
In the sportswear industry, shoes have long been assembled through labor-intensive processes involving cutting, stitching, gluing, and sewing together as many as 200 individual components. On says its LightSpray manufacturing technology reduces a shoe’s construction to seven parts while eliminating an estimated 200 to 300 hours of human work across the traditional development and production cycle.
Many manufacturers have introduced automation into existing production lines. On says it’s taken a different approach, designing a manufacturing process around robotics from the outset.
“We’re not trying to rebuild an existing style with robots,” cofounder and co-CEO Caspar Coppetti told Fortune during a tour of the company’s manufacturing lab. “We designed a technology that can only be done by robots.”
LightSpray is the latest example of an engineering-first approach that has shaped On since its founding. The company filed roughly 200 patents last year alone, spanning footwear, apparel, and manufacturing technologies. That culture of experimentation dates back to its earliest days, says cofounder and co-CEO David Allemann, when the prototype looked more like a garage experiment than the foundation of one of the world’s fastest-growing sportswear companies.
Former professional triathlete Olivier Bernhard had grown frustrated with running shoes that forced athletes to choose between cushioning and speed, so he began attaching pieces of garden hose to the bottom of an existing shoe. He showed the prototype to future co-founders Allemann and Coppetti, who saw the potential in the concept.
“It was an incredible feel,” Allemann recalled. “We started thinking about other sports innovations throughout history and realized engineering has always changed the feel of sport.”
Founded in 2010, On introduced its first shoes later that year. More than a decade later, the company has surpassed 3 billion Swiss francs in annual revenue and built a market capitalization of roughly $15 billion following its 2021 debut on the New York Stock Exchange. Footwear remains the core of the business, generating more than 90% of revenue, while the Americas, led by the U.S., account for nearly 58% of net sales. Asia-Pacific, which generates 17% of global sales, is the company’s fastest-growing market.
Its rise has unfolded during a period when Nike and Adidas continued to dominate global athletic footwear, making On one of the industry’s fastest-growing challengers. Still, its revenue remains a fraction of the legacy giants, which generate $46 billion and $29 billion, respectively.
Along the way, the company has attracted investors like Roger Federer, expanded its retail footprint, collaborated with brands such as Loewe and Post Archive Faction, and enlisted Zendaya for global campaigns. Its shoes are now as common in airport terminals and city streets as they are on marathon courses.
Engineering by iteration
Walking through On’s shoe archive, Allemann and Coppetti point to shelves lined with prototypes documenting years of experimentation. Each iteration represents another attempt to refine cushioning, stability, and energy return.
A few buildings away, On’s athlete lab measures what the eye can’t see. Around 20 high-speed cameras surround treadmills while force plates embedded in the machines record how runners strike and push off with each stride.
Athletes wear oxygen masks that measure energy expenditure at different speeds and heart rates as motion-capture software generates three-dimensional models of their legs and feet. Together, the systems help engineers evaluate how subtle changes in a shoe affect running efficiency.
The company complements those controlled tests with real-world testing alongside elite athletes. Olympic marathon champion Helen Obiri helped develop the LightSpray racing shoe alongside the On Athletics Club in Boulder. Engineers traveled there twice a month, following her workouts on bicycles, filming her stride, and asking questions about how the shoe performed.
Reinventing the factory
Developing the LightSpray racing shoe took nearly a decade. Production has expanded from Zurich to South Korea, where robotic systems are manufacturing LightSpray shoes as the company works to scale the process.
Coppetti says the technology shifts manufacturing from labor-intensive production toward robotics, a tradeoff he expects will become more economical over time while giving On greater flexibility in where it manufactures shoes.
“Basically, this is capex versus human labor. The long-term trend is definitely that it will become more economic. It also allows us to do it anywhere in the world, and we’re not dependent on a pool of people.”
One station still relies on manual work, where the outsole is attached before shoes undergo the LightSpray process. That step is still performed in Asia, though Coppetti says On is working to automate it. “There shouldn’t be any people involved here,” he says.
The technology also changes how the factory operates. Traditional production lines manufacture thousands of shoes in a single size before switching production. Allemann says the LightSpray robot can alternate between different shoe sizes, giving On greater flexibility in how it manufactures shoes.
Coppetti believes robotics will reshape footwear manufacturing over the coming years. He envisions smaller production facilities located closer to customers, shortening the distance between product development, manufacturing, and the consumer.
Fifteen years after Bernhard attached pieces of garden hose to the bottom of a running shoe, On’s sportswear technology has changed dramatically. But its co-CEOs say every experiment and product still starts with the same question: Can movement feel better?
