How AI is changing the game: can South Africa’s youth football academies keep up?

Traditionally, football training has been simple: listen to the coach, complete the drills, play the game, receive feedback, and return for the next session. This model is now being challenged, especially as football academies sell an experience, and parents judge it by results.

Imagine a young player arriving at the training pitch after school, boots in hand, ready to sharpen their passing, speed, precision, and decision-making. The coaches are watching, taking notes and providing feedback. But what if, alongside the coach, an intelligent digital system can now answer the questions every player wants answered: what happened, when and where it happened, who was involved, why it happened and how to do it better?

AI-infused training combines human expertise with data-driven intelligence, personalised feedback and performance analytics. It can show where and when a player lost possession, which passes changed a game, what the player could have done better and why, and how fast or slowly they ran across a whole session.

This is also a business question. Youth football academies do not simply sell training sessions. They sell a service experience in which young players expect to learn, improve, interact and feel that their development matters, and in which parents expect to see results for the fees they pay.

So are South Africa’s youth football academies ready to adopt AI-infused training to improve the quality of their service? And those already using it: are they keeping up with the latest trends, or still offering yesterday’s training experience to tomorrow’s footballers?

The customer is the player and the parent

In youth football, there are two customers. The player experiences the training, and the parent usually pays for it and decides whether to renew their child’s membership. If either feels the academy is not delivering, the family may not return. A young player may reasonably begin to ask:

• Where is my performance data?
  
• Why am I doing this particular exercise?
  
• How am I progressing compared with before?
  
• What should I improve before my next match?

Parents ask their own questions: is my child progressing, are they safe, and is this worth the fees? Clear, data-backed progress reports can answer all of these questions. These are not merely technology questions. They are service-quality questions, and they bear directly on satisfaction and membership renewal.

From the coach’s eye to the intelligent football environment

Football performance analysis has traditionally relied on coaches’ observations, manual data collection and standard statistics. These methods remain valuable, but they are time-consuming and can miss the full picture of an individual player’s performance. Research on youth football academies (Shi, Pyun and Manoli, 2022) points to personalised feedback, tactical analysis, injury prevention and technological engagement as important contributors to perceived service quality and player experience.

AI adds another layer. Instead of treating every player as if they have the same strengths, weaknesses and development needs, coaches can use AI to identify individual performance patterns and design more personalised training. Tools such as Hudl’s video analysis platform, Pixellot’s automated match cameras and wearable trackers offer customised training plans, performance tracking, biometric monitoring and real-time feedback. The technology is already close to home. In recent seasons, the Premier Soccer League has partnered with Hudl for video and performance data across the Premiership, the Championship and the DStv Diski Challenge in South Africa, which means data is becoming part of the professional pipeline that young players hope to enter. For smaller academies, an affordable first step is simply a smartphone on a tripod, with the clips reviewed together with the players (see the tips below). The Hudl partnership was reported by Soccer Laduma.

For a young player, training can become less about completing drills and more about understanding why those drills matter. They begin to see the game through numbers, movements, patterns and personalised insights.

A machine can identify patterns. A coach can interpret them. A system can tell a player that their acceleration has declined. A coach can ask whether the player is tired, injured, struggling emotionally or simply having a difficult day. AI generates information, but the coach provides context. For football academies in an emerging football market, the future is therefore not AI versus the coach. It is AI plus the coach.

One study by Steenbok et al. (2025) conducted in South Africa investigated youth soccer player development pathways and found that coaching, personal mindset, and social support were among the most important factors influencing player development, as well as the importance of personalised feedback and guidance. This is local evidence for the “AI plus the coach” idea: people and personalised feedback matter alongside the technology.

What stands in the way in South Africa

South Africa’s academy landscape is highly unequal. A well-funded club academy and a community academy in a township or rural area face very different realities. The barriers are practical:

• Cost: many platforms and devices are priced in foreign currencies, which is hard for small academies to absorb.
  
• Connectivity and power: cloud-based tools need reliable data and electricity. Offline-capable tools, or recording on a phone and uploading later, can help where data is costly or during load-shedding.
  
• Skills: data is only useful if coaches can interpret it and explain it to a teenager.

There are also human risks. Constant monitoring can pressurise young players, and tools built on data from other countries may not fit local players. The evidence that these tools improve performance is still developing. A 2022 pilot study with professional players (Herold et al., 2022) found a positive tendency from a data-science intervention; however, it concluded that the transfer of knowledge to actual performance was lacking. Because the players were professionals, this is indirect evidence for youth academies. Data should support development. It should never become a weapon against a young player.

Where should academies start?

a. Start small. Begin with video and basic tracking before investing in full platforms. For example, recording matches on a phone or tripod and reviewing clips with players costs far less than a full platform.
b. Train coaches to interpret data and to talk about it with players. For instance, turn one measure into one sentence a 14-year-old can act on.
c. Share simple, regular progress reports with players and parents. A monthly or termly report could cover attendance, two or three development measures and one focus area for the next period.
d. Put a consent and data-protection policy in place before collecting any data. Because the players are minors, this includes parental consent and careful handling of children’s personal information under POPIA. POPIA, South Africa’s data-protection law, sets stricter conditions for processing children’s personal information, which generally means consent from a parent or guardian, and biometric data from wearable devices counts as special personal information. Tell families what is collected, why, who sees it and how long it is kept, and check the details with a legal adviser before launching.

So, can they keep up?

Based on what we know so far, probably not yet, and not evenly. The gap is less about talent than about money, skills and trust. Academies that treat data as a service and development tool, rather than a gadget, will serve players better, keep more families and build a stronger pipeline of talent. Local research on how technology models service quality and membership renewal in football academies is still limited, which is an opportunity for future studies.

Lehlohonolo Amos Masitenyane is Senior Lecturer in the CBE at the AI-Infused Football University of Johannesburg. His research focuses on Services Marketing.

Sources

Androulakis, N., & Koundourakis, N. (2026). Adaptive athlete management in youth football: A developmentally informed framework for applied practice. Sports, 14(8), 324.

Herold, M., Kempe, M., Ruf, L., Guevara, L. & Meyer, T. (2022). Shortcomings of applying data science to improve professional football performance: Takeaways from a pilot intervention study. Frontiers in Sports and Active Living, 4. doi:10.3389/fspor.2022.1019990

Shi, Y. (2022). Measuring and achieving service quality for youth football academies in China: a perspective from players. Loughborough University. Thesis. https://doi.org/10.26174/thesis.lboro.19700167.v1

Shi, Y., Pyun, D. Y., & Manoli, A. E. (2022). Assessment of the service quality measurement model for youth football academies. Sport Marketing Quarterly, 31(4), 305-321. https://doi.org/10.32731/SMQ.314.1222.05

Steenbok, W., Morris-Eyton, H., & Kubayi, A. (2025). An investigation into the player development pathway among youth soccer players in South Africa. South African Journal of Sports Medicine, 37(1).

Soccer Laduma. (2025). PSL partners with Hudl for player data and match analysis across all divisions. https://soccerladuma.co.za/news/local/psl-partners-with-hudl-for-player-data-and-match-analysis-across-all-divisions