Digital marketing has been commoditising for as long as it has existed. Every capability that once required specialist expertise, budget, and time has eventually become accessible to anyone with a laptop and a credit card. Email marketing, social publishing, SEO, paid media, data analytics, content production: each went through the same cycle. Barrier high, then lower, then nearly gone. AI in sports is now compressing that cycle dramatically: what used to take a decade to commoditise has landed in two years.
I have been working in digital strategy and commercial growth for twenty years. The consistent pattern across all of it has been this: as the tools commoditise, what you sell shifts upward. The expertise. The judgement. Making the right calls at the right moment. Skills democratise as tools get cheaper and more accessible. What once required a specialist agency becomes something a capable in-house team can handle. What the specialist retains is the ability to know which tool, applied to which problem, in which sequence, produces the right commercial outcome.
AI has accelerated this cycle significantly. Capabilities that might have taken another decade to commoditise have landed in the last two years. And alongside the automation, AI has created a raft of entirely new tools that most organisations have not yet worked out how to use commercially.
The gap that results is not just about which tools apply to which problem, or how to configure and evaluate them. It goes further than that. The real opportunity, and the real complexity, is building on top of what already exists. Agent frameworks. Application layers. APIs that connect a club’s proprietary data, its ticketing history, its hospitality records, its fan behaviour signals, to the intelligence that AI models can provide. The frontier models have been trained on the public internet. Less than five to ten percent of the data that sits inside companies and organisations has been touched by any of them. That internal data is where the commercial value concentrates, and connecting it to AI capability is specialist work that requires both technical depth and domain knowledge.
That is where the capability gap actually sits. And in elite sport, it is significant.
Access is not the same as capability
The website example still holds, twenty years on. You can build your own website now. Most organisations probably could, at a basic level. But you would still bring in someone who knows what they are doing, because there is more to it than the initial build: managing it, evolving it, making sure it is actually doing the commercial job it is supposed to do. The tool being accessible does not mean the capability to use it well is evenly distributed.
The same pattern applies to data analysis, market research, partnership valuation, fan segmentation, and hospitality personalisation. The tools to do all of these things exist and are accessible. The constraint is the capability to frame the problem correctly, configure the tool accordingly, evaluate the output rigorously, and translate the result into a commercial decision.
Building on what is already there: agent frameworks and proprietary data
The access vs capability gap extends further than tools. The bigger opportunity, and the bigger complexity, is in the layer above the tools: building agent frameworks and application layers on top of what already exists inside the organisation.
An agent framework that connects a club’s CRM to a language model, so that the commercial team can query their own data in plain language and act on the output, is not a case of picking up a new tool. It requires an understanding of how the models work, how to structure the data for them, how to evaluate whether the outputs are accurate, and how to build the workflow around it so the commercial team can actually use it.
This matters because the frontier models (GPT, Claude, Gemini) have been trained on the public internet. Less than five to ten percent of the data that sits inside companies and organisations has been touched by any of them. That internal data (ticketing history, hospitality records, fan behaviour signals, partnership terms) is where the commercial value concentrates. Getting it to work with AI intelligence is specialist work that requires both technical depth and domain knowledge. It is not something a generic AI subscription solves.
Why sports organisations face a specific version of this challenge
Elite sport organisations have a particular set of characteristics that make the capability gap more acute than in most sectors.
The commercial operation is complex. A Premier League club or a national stadium is running hospitality, ticketing, partnerships, licensing, tours, conferences, and year-round venue hire simultaneously, often with lean commercial teams. Each of those revenue lines generates data. Very little of that data is being used to drive commercial decisions in real time, because the capability to connect the data to the decision has not been built.
The technology investment has already been made. Most elite sport organisations have CRM platforms, data warehouses, ticketing systems, and booking infrastructure. The raw material for AI commercial applications exists. What is missing is the application layer: the agent frameworks, the APIs, and the knowledge of which models to connect to which data, for which purpose, evaluated against which outcome. Building that layer requires someone who understands the technology and understands the commercial operation it is being built for.
And the pace of change means waiting carries a real cost. The organisations building this capability now are generating institutional knowledge, refining their evaluation frameworks, and accumulating the data signals that make AI applications more accurate over time. The gap between those organisations and the ones that have not started is not staying constant.
The new role of the external partner
Producing deliverables is no longer the constraint. If a commercial director at a Premier League club needs a content piece, a data report, or a market analysis, the tools to produce those things are available to their team right now. Bringing in a partner to produce the deliverable is a choice, not a necessity.
What remains genuinely valuable is the capability layer: the knowledge of how to frame the commercial problem, which tools and models apply to it, how to build the agent frameworks and data connections on top of existing infrastructure, and how to translate the output into decisions the commercial team can act on. That is specialist knowledge. It requires both a deep understanding of the technology and a deep understanding of the commercial context it is being applied to.
In elite sport, that combination is rare. Understanding AI commercial applications well is hard. Understanding the revenue pressures, data relationships, and commercial operating models of Premier League clubs and national stadiums is also hard. Doing both is what creates genuine value.
Where Earl sits in this
Earl was built around this specific gap. Not as an agency producing assets, and not as a technology platform selling tools. As a partner that understands both the AI commercial capability and the elite sport commercial context, and works inside the problem rather than alongside it.
The first step in most engagements is understanding exactly where the capability gap sits for this venue, with this commercial team, against these revenue targets. Not a generic AI audit. A structured process for mapping the specific applications with the highest commercial return, given the data and infrastructure that already exists, and identifying the agent frameworks and data connections that need to be built to reach them. The AI Visibility Diagnostic is built around that: a 4-6 week, fixed-fee process that produces a specific commercial picture and a prioritised set of next steps.
The commoditisation of digital marketing has been running for twenty years. AI has accelerated it significantly and added a new layer of capability on top. But building that new layer, connecting proprietary data to intelligent models, building the frameworks that make AI actually useful inside a commercial operation, is not commoditised. It requires skill, judgement, and domain knowledge. The question for elite sport organisations is whether they build that capability now, with a partner who understands the context, or wait until the gap to the organisations that already have it is harder to close.