In 2019, Leatherhead FC, a club at level seven of the English football pyramid in the Isthmian League, announced a partnership with IBM. The idea was to bring artificial intelligence to player recruitment and match analysis. It was a pilot: a non-league side with a thin budget would use enterprise-grade tools that had previously been reserved for top-flight clubs with dedicated analytics departments. The 2019-20 Isthmian League season was declared null and void in March 2020 because of the COVID-19 pandemic. The collaboration did not continue into subsequent seasons. Leatherhead FC returned to conventional ways of finding and assessing players. The venture yielded no measurable on-field results because the season was abandoned before the AI system could influence a single competitive match.
The arrangement was unusual. IBM, a multinational corporation headquartered in Armonk, New York, supplied the platform and expertise. Leatherhead FC, based at Fetcham Grove in Leatherhead, Surrey, supplied a real-world test environment. The Football Association classifies step 3 and below as non-league football. Leatherhead sat at step 7, far from the professional game. The question was whether AI could give a competitive edge to a club that could not afford a full-time scout, let alone a data science team.
What IBM Deployed and How It Worked
IBM adapted its existing AI platform to Leatherhead’s needs. The system was designed to ingest video footage of matches and training sessions, then identify patterns that human scouts might overlook. It analyzed player movement, pass completion rates, defensive positioning and off-the-ball runs. It also processed historical data to suggest potential transfer targets who fit the club’s playing style and budget.
The adaptation for a semi-professional outfit involved simplifying the interface and reducing the computational load. IBM engineers built a version that could run on standard laptop hardware, rather than requiring dedicated servers. This was a deliberate move to test whether intelligence tools could be democratized for lower-league football. The coaching staff received training on interpreting the outputs, though IBM personnel were still needed to set up and maintain the data pipeline.
The specific IBM products used were not disclosed in detail. The core platform was likely IBM Watson, given the company’s marketing focus at the time, but the records do not confirm this. What is known is that the system processed video data and delivered reports on performance metrics that the coaching staff had not previously tracked systematically.
How a Non-League Club Accessed Enterprise AI
A Proof of Concept, Not a Transaction
The partnership was not a commercial deal in the usual sense. Leatherhead FC did not pay market rates for IBM’s platforms and consulting time. IBM treated the engagement as a proof of concept, a chance to show that its intelligence software could work in a resource-constrained setting. The exact financial terms or sponsorship value between IBM and Leatherhead FC are not publicly known.
Time-Consuming Tasks, Automated
For Leatherhead, the benefit was hands-on access to capabilities that would otherwise have been out of reach. The club’s annual recruitment budget was a fraction of what a single Championship side might spend on a junior analyst. The system promised to automate the most time-consuming parts of scouting: watching hours of video, logging events and comparing players across different leagues and levels.
A Story Both Sides Could Tell
The initiative also carried reputational value. Leatherhead FC gained national media attention as a case study in football innovation. IBM received a narrative about accessible intelligence that it could use in marketing materials aimed at other mid-market organizations. Both sides understood the arrangement as a pilot, not a permanent fixture.
The Metrics the AI Was Built to Analyze
On-Field Actions, Quantified
The system was configured to track a specific set of on-field metrics that Leatherhead’s coaching staff considered important for their level. These included pass accuracy under pressure, successful tackles, interceptions and the distance players covered during matches. It also attempted to measure less obvious attributes: how quickly a player reacted to a turnover, if they made runs that created space for teammates, and how often they were caught out of position.
Finding Talent Beyond the Radar
The scouting target list was built from these metrics. The intelligence platform would flag players whose statistical profiles matched the club’s needs, regardless of whether those players were already on Leatherhead’s radar. In theory, this could uncover talent playing at lower levels or in other part-time leagues where traditional evaluation was impractical.
Weaknesses, Spotted Before Kickoff
The system was also used for tactical analysis. The coaching staff could review opponent formations and identify weaknesses: a full-back who tended to push forward and leave space behind, a midfielder who rarely tracked back, a striker who was vulnerable on their weaker foot. These insights were intended to shape match preparation and in-game adjustments.
The 2019-20 Isthmian League season was abandoned and expunged in March 2020 because of COVID-19. The AI system had been operational for only a few months. It had processed some match footage and generated preliminary reports. The coaching staff had begun to incorporate the outputs into their decision-making. But no competitive matches were played after the curtailment. The initiative did not resume for the following season. Leatherhead FC returned to conventional scouting and analysis methods.
It cannot be established that the intelligence tools directly led to any specific player signing or sale. If Leatherhead FC’s league position improved during the partial season, the evidence does not exist. The project produced no measurable improvement in league standing or player performance that can be attributed to the AI. The broader question of AI democratization in lower-league football remains open. The Leatherhead experiment showed that a seventh-tier club could connect with enterprise-grade platforms. It did not show if that connection yielded a competitive advantage. As of July 2024, the position since the pilot ended is not established here.
