NTT DATA Adopts SAP SuccessFactors and Joule for 232,000 Employees to Become a Sales Case

An Internal Contract with a Dual Purpose

SAP announced on August 19 that NTT DATA will replace multiple legacy human resources systems with SAP SuccessFactors integrated with SAP Business Data Cloud and Joule, the AI orchestrator from the German supplier. The scope covers approximately 232,000 employees that the Japanese group employs in over 50 countries, with 35% in India, 15.6% in Spain, and 6.3% in the United States. The migration is expected to take 12 months.

The detail that changes the reading of the contract appears in the middle of the release: NTT DATA will lead its own implementation to then sell it to clients. This is the classic mechanism of turning internal HR into a commercial reference, the same movement that Accenture made when migrating hundreds of thousands of employees to SAP over the last decade and which Capgemini repeated with SuccessFactors three years ago. What changes in 2026 is the sales argument: it is no longer about cloud versus on-premises; it’s about AI agents operating within the ERP.

What Joule Does and Where It Fits

Joule is the agentic layer that SAP has positioned as a single interface for natural language queries regarding HR, finance, and supply chain data. In the NTT DATA scenario, the promise is that a manager can ask how many certified consultants in a specific cloud are available for a quarter and receive the answer with data from SuccessFactors, the Business Data Cloud, and NTT DATA's own workforce planning system that it maintains in production. This is not augmented HR; it’s a router between fragmented databases.

The business model of Joule is what makes this contract relevant for the C-Level. Over the past few months, SAP has begun to charge execution fees per agent, in the same pay-to-play format that ServiceNow and Workday have adopted in parallel. The math changes: the cost per agentic query adds to the subscription fee of the module, which requires governance of use similar to that which FinOps teams have created for API consumption in the cloud. NTT DATA will have to instrument this in production before recommending it to clients.

How Clients Perceive This

Two readings intersect. In Madrid, where NTT DATA maintains 15.6% of its global workforce and operates nearshoring centers servicing Germany, France, and the UK, the migration is an operational test at scale with the European AI Act regime fully in effect since August 2. In Bangalore, where 35% of employees are located and where most of the managed services factory for American clients resides, the contract is the basis for the next pitch: if Joule reduces HR's administrative burden, the same model will be offered to banks, insurers, and global retail under the umbrella of AI-assisted talent management.

The Brazilian weight is modest in the global headcount, but the reading of the local market follows because banks and the utilities sector in the country already operate with SuccessFactors and will inherit Joule through contractual means, not project decision. It’s the same pattern that allowed Copilot to reach thousands of Microsoft 365 contracts last year without CIOs having explicitly chosen it.

The Weak Point of the Argument

The critical reading that needs to appear in the committee is known. A commercial reference built on the internal implementation is useful, but it does not replace third-party audits on total cost and real productivity gains. Accenture published in June earnings guidance below Bloomberg's consensus, even after two decades using its own environment as a sales argument with SAP and Oracle. The model works as both narrative and sale; the independent variable that no one has yet measured is the measurable return of the AI agent in real flow, and this is where the NTT DATA case will need to deliver data, not anecdotes, over the next 12 months.