SAP spent years asking investors to look through the disruption created by its move from licences to subscriptions. By the end of 2024, the burden of proof had shifted.
Current cloud backlog reached €18.1 billion and total cloud backlog €63.3 billion. More strikingly, SAP said about half of fourth-quarter cloud order entry included AI use cases.
Backlog made the transition measurable
Backlog is not the same as recognised revenue, but it provides visibility into contracted demand. SAP's 2024 figures showed that cloud ERP was no longer an adjacent growth business sitting beside a legacy franchise.
It was becoming the economic centre of the company, with Business AI attached to transactions customers were already making.
AI worked because SAP owned the workflow, not the model
SAP's advantage in generative AI was never likely to come from building the world's most powerful foundation model. It came from access to enterprise processes, permissions and structured business data.
Embedding AI into finance, procurement, HR and supply-chain workflows creates a commercial path that pure model vendors have to reach through partnerships. The 2024 order data suggested customers were beginning to buy that integration.
The result changed the benchmark for European enterprise software
The strategic question after 2024 was no longer whether SAP could complete a cloud transition. It was whether the company could use AI to increase wallet share without eroding trust, data sovereignty or implementation quality.
That is a more demanding test, but also a more attractive one. SAP had moved from defending an installed base to monetising the operating system of large enterprises again.