When a space company builds a lunar lander, integrates a satellite or delivers a multi-year payload program, it usually does not wait until the day of delivery to report any revenue. Instead it recognizes revenue over time, gradually, as the work is performed. The reason is an accounting test about control: when control of the asset transfers to the customer continuously over the build, rather than in a single moment at handover, the company recognizes revenue across the life of the contract in proportion to the progress made. For long-cycle space programs that can run for years, this over-time treatment is what produces a steady revenue line instead of lumpy, delivery-day spikes.

The mechanics are spelled out plainly in the filings. In its annual report on Form 10-K for the year ended December 31, 2025, filed with the U.S. Securities and Exchange Commission, Intuitive Machines describes how it accounts for contracts that bundle many tasks into a single deliverable, and how it measures progress.

"Where the customer contracts with us to provide a significant service of integrating a complex set of tasks and components into a single project or capability, those contracts are accounted for as single performance obligations. We recognize revenue generally using the cost-to-cost method, based primarily on contract costs incurred to date compared to total estimated contract costs at completion."— Intuitive Machines, Form 10-K (FY2025), source

Two ideas in that passage govern how the revenue line behaves. The first is the single performance obligation. A lunar mission or a satellite build is not sold to the customer as a pile of separable parts; it is one integrated capability. Accounting for it as a single performance obligation means the company measures and recognizes revenue against the whole job rather than booking each component as it is finished. The second is the cost-to-cost method, the practical engine of percentage-of-completion accounting: the company estimates the total cost to complete the contract, tracks the cost actually incurred to date, and recognizes that same proportion of the contract's revenue. If a program is forty percent of the way through its estimated costs, roughly forty percent of its revenue has been earned.

The estimate at the center of the method

Because the measure of progress is costs incurred against total estimated costs, the entire method hinges on an estimate that management updates as the work unfolds. Other space companies describe the same machinery in their own words. Rocket Lab, in its Form 10-K, states that for revenue recognized over time it uses "an input method, based on costs incurred relative to total estimated costs at completion to estimate the percentage of completion," and notes that estimating physical and technical progress on spacecraft "requires judgment." That judgment is unavoidable: nobody knows the final cost of a complex space build with certainty, so the percentage-of-completion figure rests on a forecast that can move.

The consequence is that revenue under this method is sensitive to changes in the cost-to-complete estimate. If a program turns out to be more expensive than expected, the total estimated cost rises, the percentage complete falls, and previously recognized revenue and margin can be revised in the current period through a cumulative catch-up adjustment. Auditors treat these estimates as a focal point precisely because of that sensitivity. The judgment involved in determining whether control transfers over time, and in measuring progress, is the kind of area filings flag as a critical accounting matter, because a small change in assumptions can move reported results.

Why it matters to a reader of the filings

It is worth distinguishing over-time recognition from its alternative, which the same filings also use. Not all space revenue is recognized over time. Where control of a product transfers to the customer at a single moment, typically on delivery or formal acceptance, revenue is recognized at that point in time instead. A satellite sold and handed over, or a piece of ground hardware accepted by the buyer, may be recognized at a point in time, while a complex, integrated build performed under a continuous transfer of control is recognized over time. The deciding test in each case is whether control passes continuously or in a single event, and companies disclose that they apply judgment to make the determination contract by contract. Rocket Lab's filing describes this as a critical audit matter precisely because the same revenue can be recognized very differently depending on "whether control transfer to a customer over time," a judgment that materially changes the timing of reported results. Two space companies doing similar work can therefore show different revenue patterns based on how their contracts are structured and how control is found to transfer, which is why the recognition policy, not just the revenue total, belongs in any careful read.

For anyone reading a space company's income statement, the over-time method changes what the revenue number means. It is not a record of products shipped; it is a measure of work performed against contracts, scaled by an internal estimate of total cost. That makes the revenue line smoother and more forward-looking than delivery-based recognition, but it also makes it dependent on the quality of the company's cost forecasting. A reader who wants to understand the durability of reported revenue should look past the headline figure to the disclosures about estimates: how the company measures progress, whether it has taken cost-growth or loss-contract charges, and how concentrated its revenue is in a few large, long-cycle programs whose estimates carry the most weight. The over-time method is standard, it is well defined, and it is honest about its own dependence on judgment. Reading it correctly means reading both the revenue and the estimates that produce it, because in this method the two are the same story.