The next trillion-dollar company will sell outcomes, not software licenses

Julien Bek has placed a specific bet on where the largest company of the AI era will come from. “The next trillion dollar company will be a software company that masquerades as a service business.” That is the call, stated plainly, and it is checkable.

The distinction Bek is drawing matters more than it might first appear. A software company sells a tool. A services firm sells an outcome, staffing the work, absorbing the complexity, and billing for results rather than licenses. What he is describing is a company that operates like the latter while being built like the former: software margins underneath, a services contract on top.

The logic for why that structure wins is straightforward. Buyers in most large industries already have established budgets for outsourced labor and third-party delivery. A company that sells only a software tool addresses a fraction of that total spend. A company that sells the outcome, and uses software to deliver it, can address the full budget the buyer has already allocated to getting the work done.

The next trillion dollar company will be a software company that masquerades as a service business. Julien Bek

There is a second pressure that pushes in the same direction. Founders building pure software tools face a specific and growing threat: each new generation of foundation model can make their product redundant by including similar functionality in the base offering. A company that sells the work, rather than assistance with the work, is positioned differently. Every improvement in the underlying model makes the service faster and cheaper to deliver, rather than threatening it. The same model capability that would commoditize a standalone tool becomes a margin expansion event for a services-positioned company.

The structural opportunity Bek is pointing at is not a new category so much as a reclassification of an existing one. Outsourced labor budgets already exist at scale. The buyer already accepts third-party delivery. What changes is that the headcount doing the work can be replaced, largely or entirely, by software running at a fraction of the cost, while the contract and the relationship remain in the services idiom the buyer already trusts. The company captures the spread between what the client pays for an outcome and what it costs in software to deliver it.

What would have to be true for this call to land? At minimum, at least one company operating in this model would need to reach a trillion-dollar valuation within a timeframe the current investment cycle makes plausible. The bet fails if enterprise procurement habits, regulatory requirements, or liability questions prevent software-delivered services from scaling in the verticals where the spend is concentrated. It also fails if foundation model providers move directly into services delivery themselves, collapsing the opportunity before any intermediary can consolidate it.

The call is large enough that it will be easy to dismiss and, if it lands, impossible to ignore. A reader tracking this bet needs to watch whether any AI-native company begins selling outcomes rather than tools in a major professional or operational vertical, and whether it compounds at the rate that trillion-dollar results require. The reasoning is coherent. The evidence for it is still being assembled in real time.