What business leaders should learn from the Tour de France and the World Cup

Right now, at the World Cup, cameras track every player's movement and generate more than 150 million data points a match, while a sensor inside the ball fires 500 times a second. An AI system condenses all of it in real time, flagging a likely offside for the referees within seconds, instead of the minutes of replays and arguing it used to take. At the Tour de France, teams are turning years of data into predictive models, all to claw back the few seconds that decide a mountain stage.

Two of the biggest sporting events on the planet, both now powered by AI. In sport, it raises a real debate. Is AI cheating, or a fair competitive edge? Some things stay the same whether you're running a team or a company, but in business nobody asks that question. The only one that matters is whether you can make AI actually work, and on that, these teams have plenty to teach. If your job is turning AI proofs of concept into systems that run in production, this one's for you.

The gap isn't your data, it's your decision loop

Watch how a modern offside call actually gets made. It runs on 16 cameras tracking 29 points on every players body, a 3D digital twin that captures each player's exact proportions, and a sensor in the ball that marks the precise instant the pass is made. What's impressive isn't the volume of data, it's that all of it converges into one clear call the officials can act on the moment the play happens.

Most companies already have that same advantage sitting in their systems, with real-time data flowing in from operations, sales and logistics. The trouble is that it tends to wait there while the decisions are made on last weeks reports. The value isn't in collecting more data, it's in closing the distance between what you already have and the moment someone acts on it, with real-time pipelines that put the answer in front of the decision-maker while the decision still matters.

Nobody wins on one breakthrough, they win on gains that stack

With today's super teams like UAE Team Emirates, you'd assume the Tour de France is won by a heroic move from the leader. In reality, it's won on marginal gains that compound over three weeks, most of them too small to notice on their own: nutrition, aerodynamics, recovery, tire pressure, an extra hour of sleep. Stacked together, they decide who ends up on the podium, which is why cycling has quietly become one of the most data-driven sports on earth. UAE Team Emirates run their own AI platform that turns power output, heart rate, sleep and blood glucose into a daily plan for each rider, and their biggest rival, Visma-Lease a Bike, has teamed up with AI company Mistral to keep pace.

Enterprise AI rewards the same attention to detail. Sometimes the return is a full rethink of how you work, top to bottom. Other times the biggest gains hide in the smallest details, dozens of optimizations that compound quietly until they move something as big as market share.

When Aeromag built AI into its aircraft deicing, it shaved about a minute off each deicing operation. Across a full season, that adds up to roughly 130,000 minutes of optimized ground time and more than a million dollars in annual savings for its airline partners, with less fuel and fewer emissions along the way. A minute at a time, across an entire operation.

The best AI keeps a human in the loop

Here's the part people tend to miss when they worry about handing decisions to a machine. The offside system never blows the whistle. It flags a possible call and alerts the officials, but a human still makes the decision. It works the same way in the peloton, where the AI recommends a plan and the director decides what the team actually does. That division of labor is the whole point, and as an AI researcher has noted, the World Cup is a clean lesson in it: let the system do what it does faster and more precisely than any human could, and keep the judgment where it belongs.

The AI that gets deployed in serious organizations works the same way. Instead of replacing people, it sharpens their judgment and hands the decision back, surfacing the answer and leaving an accountable human to make the call.

The win everyone sees is built on the work nobody does

The rider crossing the line in July is the visible part, maybe one percent of the whole effort. Behind that moment sits a year of power files, recovery data and modelling that never makes the highlight reel, so a victory that looks like a single afternoon was really built in an off-season nobody was watching.

A lot of companies want the July win without building that foundation. They want the dashboard, the prediction and the automated decision without the data infrastructure underneath that makes any of it trustworthy, and often without having asked the harder question first: what business problem are we actually trying to solve? The uncomfortable truth is that everyone can buy the same off-the-shelf tools, so the real advantage comes from pairing discipline and infrastructure with AI built around your actual problem or goal, rather than a generic product you hope will fit. That's the case for going custom, and it's the unglamorous part of the work that separates the organizations getting real returns from the ones still running pilots.

The takeaway

Elite sport stopped debating AI a while ago. It's not just soccer and cycling: baseball has been rebuilt around Statcast tracking data, and Formula 1 teams make race calls off live telemetry. AI works as an ROI engine with a human at the wheel, and the lesson for the rest of us is that the advantage is what you choose to build around it.

Key takeaways

• The bottleneck is rarely your data, it's how fast that data reaches the decision.
• AI ROI often compounds through small gains that stack, not just one big transformation.
• The AI that gets deployed keeps a human making the final call.
• The win everyone sees rests on the data infrastructure most organizations skip, and the problems you decide to solve.

If you're wondering where your own marginal gains are hiding, that's a conversation worth having.