From the outside, a sports federation’s website looks simple. Fixtures, results, standings, news, a link to the stream. A supporter finds the score. A club finds the entry form. Done.
From the inside, that page is the visible tip of an estate most federations have never mapped. We have run discovery inside an international federation and delivery for top-flight clubs in England and Saudi Arabia, and the pattern repeats at every level of the game: the website is the smallest system in the building. It is also where every other system’s problems become public.
Count the systems behind one fixture
Take a single match in a continental competition and trace what has to agree for the website to be right.
The competition structure lives in one system. Live match data – scores, cards, disciplinaries – is entered by officials into another. The officials themselves are scheduled in a third, a standalone tool that talks to nothing. The stream sits on an OTT platform with its own competition records, rebuilt by hand from the first system. Invoicing for the host, the sponsors and the entry fees runs through finance software. Sponsor commitments live in a contract on someone’s laptop. And the website takes feeds, or more often copy-and-paste, from most of the above.
In the federations we’ve worked with, dozens of competitions a season ran across this kind of estate. None of the systems shared data. When a team withdrew or a match moved – which happens every season, for reasons ranging from weather to geopolitics – one change had to be re-keyed into four or five places by three different people. Every update was a race between the systems, and the website showed whichever version lost.
The uncomfortable conclusion? The federation did not have a website problem. It had a single-source-of-truth problem. The website was just where the ambiguity went public.
Clubs have the same disease with different symptoms
A club platform looks more unified because one brand wraps it. Underneath, it is a supply chain.
On a recent top-flight club build, the supporter-facing experience – one website, one app, two languages – depended on a headless CMS from one vendor, live match data from a sports-data provider, membership through a white-label partner, commerce punched out to a marketplace, single sign-on as a separate conversation entirely, and app-store review boards with their own timelines and their own opinions.
Each of those pieces is a vendor with its own release cycle, its own outage window and its own definition of done. The club buys one experience. It operates a dozen relationships. And when something breaks on matchday, it almost never breaks inside a system. It breaks between two systems that agreed yesterday and disagree today.
When the seam fails, website skills don’t fix it
One incident makes the point better than theory.
A client of ours runs a platform with strict regional hosting requirements. During a regional cloud outage, their production server didn’t just go down – it became unreachable, with the provider’s own recovery actions unavailable. The textbook answer, fail over to another geography, was off the table: the data had to stay where it was.
The fix was nowhere near the website. It was reassigning a static IP to a standby node, reconfiguring the backend and the web server, and issuing a new TLS certificate – restoring service without a DNS change, so the audience never had to wait for the internet to catch up. Spotted by us, mitigated by us, written up with the structural recommendations that stop it happening twice. Because our team spans continents, someone senior was awake and had access when it mattered.
None of that is “website work” as the industry sells it. All of it is what keeping a website up actually requires.
What good looks like
The organisations that handle this well don’t start with a redesign. They start with three governance questions that sound almost too simple:
Where does truth live? For every fixture, result, official, venue and sponsor commitment – one system holds the definitive record, and everything else reads from it. If two systems can both be “right”, the website will eventually publish the wrong one.
Who can change it? A postponement should be one person’s decision entered once, not a phone call that triggers four manual updates. Authority and entry point should be the same place.
How does a change reach every channel? By API and automation where possible; by a named, rehearsed handover where not. The measure of the architecture is the time between a decision and every public surface agreeing with it.
Then rehearse the failure. What does the supporter see when the data feed is late? Who decides to roll back a release an hour before kick-off? Which seams have a fallback and which are single points of failure? These questions cost a workshop. Discovering the answers live, mid-tournament, costs considerably more.
The honest scope
We don’t arrive proposing to replace everything – federations and clubs carry systems for good reasons, including some, like international governing-body platforms, that aren’t theirs to change. The work is mapping the estate, deciding where truth lives, and making the systems you keep agree with each other. The website gets rebuilt when the foundations deserve it, not as a substitute for them.
Because here is the thing about the smallest system in the building: when it is fed by one source of truth, it becomes the easiest thing you run. The fixture moves, every channel updates, the supporter never knows there was a change to make. A good matchday platform is uneventful – and uneventful is engineered.