Building an asset, not a content business

There are two very different things a company in this space can set out to build. We were clear from the start about which one Axia would be, and almost every decision since has followed from that choice.

Two ways to build in this space

Walk through the sports prediction industry and most of what you find is a content business.

A content business sells opinion and attention. It produces a steady stream of confident takes, competes for clicks, follows and views, and treats reach as its core measure of success. There is nothing dishonest about that model in itself, and some of it is genuinely entertaining. But it has one defining property that shapes everything else: it has to start over every single day. Yesterday's content has almost no residual value. The audience is only ever as engaged as this morning's post made them. Stop producing for a week and the business effectively stops existing for that week.

The other thing you can set out to build is an asset. An asset does not start over each day. It accumulates. It holds value even in the periods when you are not actively promoting it, because its value is stored in something durable rather than in the freshness of the latest output. Building an asset is slower, less immediately visible, and harder to point at in the early days. It is also the only one of the two that is genuinely worth owning at the end.

Axia was set up, deliberately and from the first decision, to build the second thing. This article is about what that means in practice, because the distinction is abstract until you make it concrete.

The defining weakness of a content business

It is worth dwelling on the weakness of the content model, because it is easy to miss when the content is good and the audience is growing.

A content business is, in financial terms, running to stand still. Each day's reach has to be re-earned the next day with new output. The audience does not compound; it churns, and the business spends most of its energy replacing the attention it lost since yesterday. Its key metrics, views and follows, are real but shallow, because they measure attention rather than anything that persists. Attention is rented, never owned. The moment the content stops, the rental ends.

This produces a second, subtler problem. Because a content business lives on attention, it is structurally pulled toward whatever wins attention, and what wins attention is rarely the same as what is accurate. Confidence outperforms calibration. A bold call outperforms a measured one. A loud reaction outperforms a careful analysis. Over time the incentives of a pure content business pull it away from being right and toward being engaging, and those are not the same destination.

None of this makes a content business worthless. It makes it fragile, and it makes it shallow, and those two properties are exactly what an investor, or a founder thinking in decades rather than quarters, should care about.

What makes something an asset

If a content business fails the test of persistence, what passes it? The test we apply at Axia is deliberately simple. If the company went quiet for a month, produced nothing, promoted nothing, what would still be worth something when it came back?

For a pure content business the honest answer is: very little. The takes would be stale, the feed would be cold, and the attention would have moved on to whoever kept posting.

For Axia the answer is concrete, and it is the whole point of how the company is built. If Axia went quiet for a month, the model would still exist, and it would still have whatever edge it has, because that edge is a property of how the model is built and not of how loudly it is talked about. The published record would still be there, one month longer if anything, every prior result settled and permanent. And the credibility attached to that record, earned over every prior season, would be entirely intact. None of the valuable things would have decayed.

That is the difference between an asset and an output. An output is consumed and then gone. An asset persists, and can even appreciate, while you are not touching it. Axia set itself the harder job of building something that passes the quiet-month test, and the three sections that follow are the three things that let it pass.

The asset, part one: the model

The first part of the asset is the model itself.

The Axia Model is a genuine piece of intellectual property. It is a pricing engine: it takes the inputs that move football matches and produces its own independent probability for each market, then identifies where those probabilities differ enough from the market's prices to be worth acting on. Building it was real, cumulative work, and replicating it would be real, cumulative work too. It is not a format or a posture that a competitor can copy by watching the output. It is a system, and the system is the property.

A model of this kind has a quality that pure content can never have: it improves rather than expires. Each season of data is something the approach can be tested and refined against. The model that priced the 2025-26 season is more developed than the one that came before it, and the one that prices the next season will be more developed still. An output gets older. A model gets better. We describe how it works in our model.

The asset, part two: the record

The second part of the asset is the published record, and it is, in some ways, the most defensible thing the company owns.

The record is every recommendation the model has made, logged before kick-off and settled against the result, kept in full, including the losing months. We make the full argument for keeping it that way in tracked, not claimed. But the point here is narrower and concerns what kind of asset a record is.

A record cannot be created quickly, by anyone, at any price. That is its defining and most valuable feature. You cannot buy a two-year tracked record. You cannot hire one, or design one, or accelerate one. The only way to possess a genuine record of being right, over time, at a measurable rate, is to have been right, over time, at a measurable rate, and to have committed every call honestly before the result was known. A record is pure accumulated time, and accumulated time is the one input a competitor with more money cannot simply buy more of.

That makes the record the rarest component of the asset. The model could, in principle, be matched by a sufficiently capable competitor. The record cannot be matched at all except by waiting exactly as long as Axia has waited, under exactly the same discipline.

The asset, part three: the product

The third part of the asset is the product that the first two support: a subscription that turns the model and the record into recurring revenue.

It matters that the product sits on top of the asset rather than in place of it. Many businesses in this space are a product with no real asset underneath, a subscription wrapped around opinion. Axia is the other way around. The model and the record are the substance, and the subscription is the mechanism that converts that substance into a durable revenue line. Because the revenue is recurring, it is predictable, and predictable revenue can be planned and reinvested against rather than chased. We set out the economics of that in our products.

The three parts reinforce one another in a closed loop. The model produces the analysis. The analysis, settled honestly, becomes the record. The record is what makes the subscription worth paying for. And the subscription revenue funds the continued development of the model. Each part feeds the next, which is what makes the whole thing an asset rather than three separate things.

Why the asset compounds

The reason this structure matters, more than any single feature of it, is that it compounds. A content business cannot.

Consider what happens with the simple passage of time. Every match week, the model produces more analysis, and the record gets one week longer. But the record does not merely get longer. It gets harder to argue with, because a longer track record is statistically more meaningful than a short one. The same is true of the model, which has another week of data to be tested and refined against, and of the credibility, which deepens with every honestly settled result.

A competitor cannot shortcut any of that. They cannot buy a longer record. They cannot fast-forward the credibility. To stand where Axia stands in two years, a competitor starting today would have to run, honestly and publicly, for two years. Time, in other words, works for Axia in a way it simply does not work for a content business.

A content business is running to stand still: it replaces yesterday's attention with today's and ends each year roughly where it began, only tired. Axia is accumulating. The model keeps improving, the record keeps lengthening, the credibility keeps deepening, and the recurring-revenue product sits on top of an asset that is larger and more defensible at the end of every season than it was at the start of it. The same year that leaves a content business where it began leaves Axia meaningfully further ahead.

Why this is defensible

It is worth naming, plainly, why this structure is defensible, because defensibility is what turns an asset into a valuable one.

A business is defensible when it is hard for a competitor to replicate what makes it work. Axia's asset is hard to replicate on all three counts, and for three different reasons. The model is hard to replicate because it is genuine, cumulative intellectual property rather than a copyable format. The record is impossible to replicate quickly at all, because it is made of time, and time cannot be bought. And the credibility that sits on top of the record can only be earned, never purchased, because the moment it is purchased it is no longer credibility.

Stack those three together and you have something a well-funded competitor cannot simply outspend. They can match the marketing budget. They cannot buy back the years. That is the most durable kind of advantage a company in this space can have, and it is the direct result of choosing, at the start, to build an asset rather than a content business.

An asset is easier to value

There is one more consequence of the asset structure worth naming, and it speaks directly to a company preparing for investment.

An asset business is easier to value than a content business, and easier to underwrite with confidence. The reason is straightforward. A content business is valued on attention metrics that are volatile and that can fade quickly, which makes any projection of its future fragile. An asset business can be assessed on things that are concrete and durable: a model that is genuine intellectual property, a published record that can be inspected and reconciled line by line, and a recurring-revenue product whose retention can be measured rather than guessed.

Each of those is something a serious investor can examine directly. The record in particular is unusual in how inspectable it is. Because every recommendation was committed before the event and settled honestly afterwards, the track record is not a marketing claim to be taken on trust. It is a dataset to be audited. An investor does not have to believe Axia about the model's performance. They can check it.

That inspectability is itself part of the asset. A business whose core claim can be verified from its own records is a lower-risk proposition than one whose core claim rests on assertion, and a lower-risk proposition is, by definition, a more valuable one. Building an asset rather than a content business does not only make the company more durable. It makes it more legible to the people deciding whether to back it.

What we are building toward

The goal has not changed since the first day. Build a model with a real, measurable edge. Prove it with a record kept honestly and in the open, the losing months included. Turn that model and that record into a subscription product, in a market large enough and growing fast enough to matter. Then let time do the part of the work that only time can do, which is to compound the asset season after season into something a competitor cannot catch.

It is the slower way to build in this industry. In the early days it is less visible than a content business, because an asset under construction does not generate the noise that attention metrics reward. But it is the only way that ends with something the company genuinely owns, rather than an audience it is permanently re-renting. We would rather own an asset than rent an audience. Everything on this site, the model, the record, the products and the raise, is downstream of that one decision, and it is the decision we would make again.