What Is RPO? Remaining Performance Obligations, Explained

Remaining performance obligations (RPO) is an accounting figure that measures revenue a company has contracted to deliver but has not yet recognised: signed orders that have not yet turned into sales. For cloud computing companies it has become the market’s favourite gauge of future demand, because multi-year AI and cloud contracts are signed years before the revenue appears in the income statement.

How RPO works

Under accounting standards, when a customer signs a five-year, $5 billion cloud contract, the provider cannot book that as revenue immediately. It recognises revenue as the service is delivered, and the unearned remainder sits in RPO. The figure splits into current RPO, expected to convert to revenue within twelve months, and the longer-dated balance. A surging RPO with flat current RPO means the new contracts are back-loaded: the demand is real but distant, and it often requires heavy capital spending today to serve.

Why investors watch it, and where it misleads

RPO grew popular because it leads reported revenue, sometimes by years. But it has sharp edges. It says nothing about margin: a contract to resell someone else’s computing capacity books the same way as high-margin software. It is concentrated: a handful of AI customers can dominate the balance. It can be cancelled or renegotiated in ways footnotes only partially reveal. And converting it into revenue may depend on data centres and chips that do not yet exist. A large backlog is a promise about the future that still has to be built and financed.

What to watch

The ratio of current RPO growth to total RPO growth (back-loading), capital expenditure guidance rising alongside backlog (the cost of serving it), customer concentration disclosures, and whether reported revenue growth eventually tracks what the backlog implied. When backlog growth and capex both explode while near-term revenue guidance barely moves, the market is being asked to fund the build-out on faith.