'Data workers' behind the 2026 World Cup
The ongoing FIFA World Cup uses sensor-embedded balls, real-time tracking systems and an AI assistant for each team. Behind those innovations is the data workforce in many regions such as India and the Philippines. They are often football players or people with a deep understanding of this sport who want to earn extra income.
“The data value chain is geographical: high-value data analysis work is located in a few wealthy centers, while data annotation is concentrated in Eastern Europe, Africa, South Asia and Southeast Asia,” Rafael Grohmann, assistant professor of communication studies at the University of Toronto, told Rest of World.
Football started using data more than two decades ago and now almost every national team or major club uses data for recruitment, training, tactics, management and many other purposes. Football data also serves "derivative" industries such as television and video games.
Grohmann, who is mapping the workforce across the soccer data value chain, said labeling teams are typically based in cities like Manila, Cairo, Chennai and Ternopil. They are hired through an intermediary company, spending 3-4 hours turning raw video from the match with each pass and shot into structured data that can be analyzed.
According to a player who spent a year annotating data at Packing Sports - the Manila-based unit of German analytics company Impect, this is a popular side hustle for many Filipino players looking to earn extra income. He requested anonymity because he is not authorized to speak to the media.
This player said he watches European matches and tags every player's actions. In major tournaments such as the FIFA World Cup and UEFA European Championship, "the workload is heavier because the need for fast data from teams, analysts and media is higher".
According to Scott Powers, assistant professor of sports analysis at Rice University, American investors are paying great attention to football and own many clubs in the English Premier League and Serie A. This is likely to promote the expansion of the data processing industry in sports.
The World Bank estimates there are 154-435 million online freelancers worldwide, accounting for between 4.4% and 12.5% of the global workforce. A significant portion of them are providing data to artificial intelligence systems. As there is more demand, the workforce for labeling, cleaning, and verifying training data will grow.
World Cup data labelers earn about $70 per match, much higher than the industry average, according to Metaintro. When Time investigated the workers behind ChatGPT, they found that data annotators in Kenya working for an intermediary company received only $1.32 to $2 per hour, while customers paid the company about $12.50 for the same work.
In sports, data needs to be generated in real time. Combining computer vision and AI models, one worker can record up to 3,000 actions in each football match. "This is data that requires specialized skills to analyze," Powers explains, saying he has read job postings for sports analytics, performance science, data engineering positions from 25 of the 30 clubs on Forbes' list.
A freelance data annotator in Rio de Janeiro, who tags local football matches for a foreign sports data company, said he was paid about $70 per match before allowances, and revealed the constant updates for prediction sites.
"Prediction platforms need real-time data to make adjustments throughout the match without having to wait for official reports, which can sometimes take several days. So I can't submit data late, otherwise I risk losing part or all of my pay," this person said.
The 2026 World Cup is estimated to generate about 9 billion USD, making it the most lucrative sports tournament ever. "The football we watch is as much about the work of data labelers as it is about the players," Grohmann said.
Nam Nguyen compiled