The AI Wave: building got cheap, waiting got expensive

Every club has had the idea. Almost none have had the team to build it.

Fan engagement platform. Smarter ticketing. Real-time training load monitoring. Recruitment data that goes beyond the scout’s notebook. The ideas aren’t the problem — they’ve never been the problem. For most of the software era, one fact blocked them all: engineering talent was scarce and expensive. That scarcity justified concentrating it at the top of the sport. The clubs with the largest budgets could afford the largest tech teams. Everyone else waited in a queue that never seemed to move.

AI removed the cost. For club leaders without a deep engineering team, this is the number that changes everything. In 2023, the best AI systems could complete roughly 2% of verified, real-world engineering tasks. By 2026 they clear 80%+ — and frontier systems with agentic tooling go higher still.¹ Around 41% of new code is now AI-written, and the large majority of professional developers use AI tools every day.² People who have never called themselves engineers now ship production software; industry analysts find citizen developers already outnumber professional ones.³

The result: work that used to take a team and a year can ship in a focused few weeks.

What got cheap — and what didn’t.

Building got cheap. Waiting didn’t. Queues, approval cycles, and “next quarter” cost exactly what they always did — and in a faster-moving market, more. That’s the inversion in one line: the price of building fell; the price of delay didn’t. The rational move flips with it — stop hoarding the ability to build, and start removing the reasons you wait.

This is why small, AI-augmented teams now do things that used to require an army. Think about what that means for a club. The fan engagement app that once needed months of agency work and a substantial contract can now be scoped, designed, and shipped by a focused team in weeks. The ticketing intelligence tool that previously lived only at the enterprise end of a commercial supplier’s catalogue is now buildable in-house. A handful of software companies have already reached tens of millions in revenue with teams you could fit in a room.⁴ And it isn’t just elite startups: a large majority of small businesses have now invested in AI tools, with adoption climbing fast.⁵ The capability is no longer reserved for organisations that can afford a big engineering department.

The part the hype skips.

Here’s where we break from the breathless version — and this matters especially in sport, where poor data decisions can cost you points, players, and revenue simultaneously. The same research that proves the wave also proves it’s dangerous in the wrong hands:

• The leading delivery-performance study finds AI adoption raises throughput but correlates with lower delivery stability — its blunt conclusion: AI amplifies what’s already there.⁶
• In a randomised study, experienced developers were 19% slower using AI on mature code they knew well — while believing they were ~20% faster.⁷
• Independent security testing found ~45% of AI-generated code failed security checks.⁸
• Duplicated code has grown sharply, and the most-cited developer frustration is AI output that’s “almost right” — the kind of wrong that’s expensive to find.⁹

None of this means “don’t.” It means don’t do it without discipline. AI makes a disciplined team dramatically faster and an undisciplined one dramatically more dangerous.

What it means for your club.

For years, the tools that gave elite clubs their edge were locked behind budgets that most organisations will never see. Three market examples show what became possible at scale: City Football Group (market example) built an enterprise-grade data operation spanning multiple clubs and leagues. The Golden State Warriors (market example) built what they call a “digital brain” on Gemini and BigQuery. Brentford (market example) turned statistical analysis into a ladder from the Championship to the Premier League, competing against clubs with far greater resources. The gap felt permanent. It wasn’t — it was a function of cost, and cost has changed.

If your club doesn’t have a deep engineering team, this is the best opportunity in a decade — and the easiest way to get hurt. The fan app you’ve been discussing for three seasons, the injury analytics you know you need, the recruitment data layer you’ve watched better-resourced clubs deploy — these are no longer exclusively theirs. The capabilities are within reach. The failure modes are too. AI doesn’t fix a bad data foundation; it amplifies it. A club running on disconnected systems and untrusted numbers doesn’t get smarter by adding AI — it gets faster at being wrong. The differentiator is no longer access to AI; it’s the judgment to aim it at the right problems and the discipline to ship it safely.

That’s the work we do with sports organisations. We find the right problems first — the ones with genuine performance or commercial return — then design and build the thing. Fast enough to matter, disciplined enough to trust. The wave is real. We help clubs ride it without wiping out.