The summer transfer window is an exciting time for everyone in football. Players come and go while general managers and sporting directors hustle hard to assemble a powerful squad for the new season. Usually, the majority of teams are facing huge turnover, many times the head coach gets replaced and sometimes, even sporting directors or other executives. With every change in the management, new philosophies and fresh bold ambitions are drenched into the club. Consistency is really the rare exception and often only achieved by elite clubs or clubs with a historically entrenched philosophy. Around every corner, clubs turn things upside down and aim for significant change to finally — and, this time, certainly — achieve their goals in the upcoming season.
However, if you take a closer look, this seems to be a misconception. As mentioned, consistency often appears to be the real driver of success. Of course, leaving faulty things as they are and not fixing existing issues will not bring you anywhere but running around with your head cut off, doing things for the sake of doing things is an equally dangerous approach as just ignoring the problems that exist.
So, the problem is obviously not the involvement of the people but rather detecting the right things to work on, doing them right, and deeply understanding the origin of the realised outcomes. “If you can’t measure it, you can’t improve it” is an often-cited quote, almost old-fashioned, but it entails a fundamental truth. If you don’t understand the mechanics of your business, then you can hardly work on it efficiently. With a lack of insight, you don’t know if good or bad results are driven by your action or mere coincidence.
To improve their success in the transfer market, many clubs discovered the value of data scouting. It’s a great way to filter the vast amount of potential signings and make the whole process a bit more efficient. At first, when data was a novel emergence in football, the clubs with access to data were able to outperform their peers who were not using data yet with relative ease. Today, data itself became a commodity in football, and now it’s much more about deeply understanding the data and applying your individual models and techniques to it in order to draw the right conclusions — here we are, back at understanding the mechanics. In this sense, understanding the mechanics of your team and the way you want to play football. This is already a promising development to professionalise the club’s operations and improve the daily work. However, there is no reason to stop in this niche. While clubs do see the value of being more data-driven in their transfer policy, many clubs seem to lack the vision for applying similar approaches to their organisational and business-side operations.
The summer break offers a brilliant opportunity to circle back to your goals from the beginning of the season and check what you achieved and what you have missed. Not only on the pitch but also off the pitch. Football clubs are organisations like any other business; to avoid stagnation, the organisational side also needs goals or any equivalent management framework to grow and develop.
To flourish in modern times, organisations (including football clubs) not only need to increase their return on investment but also their return on information. Being data-driven is key to keeping up with the competition and overtaking.
The following section of this article aims to provide a kind of brief introduction to data-driven processing, while the latter part offers a practical “how-to” guide for building a data infrastructure in football clubs that fosters smart decision-making.
What does it mean to be data-driven and why is it the foundation for smart decision-making?
Being data-driven generally describes a decision-making process that involves collecting data, analysing it to extract patterns and facts from that, to finally utilise those facts to make conclusions that impact your decision-making. These are the main ingredients to create factual discussion cultures. By following this process, the decision-making of an organisation is based on actual data rather than intuition, gut feeling, or the apparent experience of a few people. Without data or with bad data processing, biases, deceptions, and flawed assumptions may intrude judgements and lead to poor decision-making.
Data-driven decision-making can be seen as a six-step process it’s rather a loop than a rigid process. First of all, the right understanding of the decision-making process (read loop) is mandatory. This means that all involved people should agree to the deployed processes to get the maximum trust and commitment among all stakeholders.
Personal vanities, know-it-alls, and the urge to distinguish oneself are harmful to any decision-making process. Once the club’s people share the same vision, they can define the missions. In this decision-making context, the missions can be understood as the questions that need to be answered within the club — now and in the future. So, the first step for implementing a data-driven culture is to develop the missions and to get everyone on board as well as define the appropriate KPIs to monitor the journey.
Within the second step, the whole thing gets a little more technical. Now, it is time to identify the data sources in the club. Data can appear in various forms with different volumes, varieties, and velocities. Depending on the project’s scope, a club could, for example, start examining just a single department’s data like marketing data or performance or scouting data. It’s astonishing how much data is already produced within an organisation.
The third step is dedicated to data cleansing and organising it efficiently. Here, the raw data is connected, centralised, and prepared for the actual magic. This step requires a bit more technical depth but today’s software landscape is broad and galloping towards continuously improving user-friendliness and the no-code area is developing at a rapid pace. These trends are democratizing the software world and tools are becoming cheaper and much easier to use.
The fourth step is about analysing the raw data and performing calculations on those numbers. You can utilise any kind from simple to sophisticated statistical models or even machine learning approaches. Once you have set up a solid raw data foundation, your infrastructure is ready for any kind of analytics to be built on top of it. In many, many cases, simply having all the numbers calculated consistently and being available centrally at any time is already a major improvement and by far sufficient for a football club’s regular use cases. Besides performing math, the translation of numbers into insights is another important aspect to consider.
Here, the fifth step comes into play — the data visualisation. This can be done in the form of interactive dashboards, which are automatically refreshed (especially useful for regularly used KPIs) or based on ad hoc requests with brief presentations or reports. The information always needs to be digestible for the audience and communicated understandably. There are many tools out there that help to create visually compelling and easy to grasp dashboards and reports. Sometimes, a plain spreadsheet can already work wonders if filled with relevant and consistent data. Here, it is really important to meet people where they are and not to come directly around the corner with super complex analyses. Rather, a common ground should be created first, on the basis of which it is possible to build and go into further detail and complexity. The presented data must be meaningful, digestible and thought-provoking to the audience. If you miss out on these points, the danger of losing trust in the data-driven approach is immense, and having stakeholders that lost trust in your data is really the worst that could happen to any data project. It’s almost the nail in your initiative’s coffin. Trust is the real currency of data.
The sixth step is finally about drawing conclusions. Here, you can merge your qualitative knowledge and experience with quantitative data to put old assumptions to test and formulate and falsify hypotheses. Do not only trust the data blindly but use it as a common foundation to build passionate and factual discussions upon it.
How to implement data-driven structures in a club?
Often enough, clubs have no single source of truth when it comes to data. Scouts and coaches handle their own Excel spreadsheets, from time to time a database provided by the FA is queried opportunistically by the video analyst and performance reports or medical data are sent by email. Sometimes, there is a subscription with a scouting platform provider but no solution for integrating the data with other existing sources. This absence of smart data management has several impacts:
- Inconsistent KPIs due to missing central logic
- No possibility to analyse data across sources
- Data silos, no data enrichment
- Limited use of data because of cumbersome manual processes
- No objective data foundation for meetings and discussions
- Disastrously expensive mistakes
With the establishment of a dedicated data department or at least appointing a single person with some technical background and understanding of data management, a whole new data management structure can be introduced with fewer resources than one may think.
Disclaimer: Now it gets a bit technical but as mentioned, this is supposed to be the practical part of this article.
With somebody owning the journey, the club can start to literally centralise all their data into a virtual layer by exploiting a technology called data virtualization combined with automated ETL processes for higher performance. By virtualizing the data a club can access all the data in real-time and deploy their logics on it consistently. The virtual layer connects and centralises any data source and enables SQL-based transformations and calculations. Based on the raw data, several core views can be easily prepared to virtually join and cluster the data into logical blocks or units with respect to its purposes. Now it is easy to distribute the data to the right consumers and/or build analytics on top of consistent metrics since any kind of front-end, like BI-tools or even Excel/Google Sheets, can be connected in seconds. The virtual layer also acts as a governance tool and guarantees that sensitive data is protected against unauthorised access.
Frequently, the heavily improved accessibility alone already leads to cultural changes within the whole organisation. Data moves into the centre of the club and its departments simply because it is available now, and people are genuinely curious to take advantage. For instance, the scouting could now take into account data from all internal and external sources, resulting in complete data sets, generating actionable insights and 360° views on all potential signings as well as the own players from the academy to the first team. The marketing team can steer their activities much more efficiently since they now know exactly how each campaign performs and what any Euro spent on marketing returns. Even the training sessions on the pitch are better coordinated because the coaching staff has the most recent tracking and performance data as well as information from the medical department right at their fingertips in an intuitive dashboard. This helps, on the one hand, to field the best starting eleven at the weekend as well as, on the other hand, to prevent muscle injuries due to fatigue. Better monitoring can also help to show the players their improvement path and to warrant the best possible training.
All that is needed to achieve such a high-performance data landscape is a clear vision in the first place, a finger twist of determination and investment in the amount of probably less than the star player’s weekly salary into technology and a person who knows a little bit about data management with decent SQL skills (potentially 1 FTE is enough to operate the system, depending on the demand and the number of data use cases, more can be expedient). That being said, it’s still a bumpy road and data projects heavily rely on the commitment of all involved people. Apart from the technical environment, the data literacy of the team is crucial for such initiatives, never skimp on this end.
If things go well, clubs are able to massively leverage their Return on Information. Often, there is actually not even a need to add more data, it is just the combination and exploitation of already existing data that skyrockets the productivity within the club and heavily increases the share of good decisions made.
In environments with such high stakes, even a small improvement in the decision-making can have a significant impact because there is so much leverage in such systems. Compared with the expenses in the first team, which are certainly valid investments and needed for success on the pitch, the organisation as such should not be forgotten. Since every ordinary club is totally focused on the first team, there is a lot of leverage on the other side of the game, and one Euro invested in the development of the organisation has often a bigger impact than a single additional Euro for the first team in comparison.
Therefore, every club should be encouraged to allocate resources wisely during the summer break and not forget the other sides of the story. Paired with strong leadership and a coherent management framework, improved data ops can be a real game-changer, independent from the level of play.
