SAP's Shift to Use-Based AI Pricing: What Every Enterprise Buyer Must Understand Before July 2026

SAP's traditional licensing model was fundamentally predictable: you bought named user licences, paid annual maintenance of 22%, and your SAP cost base was largely fixed. SAP's shift to use-based AI pricing breaks this model. For the first time, your SAP spend can grow — automatically, without any procurement decision — simply as a consequence of your users doing their jobs. If they interact with Joule, trigger AI-powered workflows, or use embedded AI features at above-forecast rates, you generate AI Unit overages that appear on your next invoice. SAP designed this model to grow its revenue. Your job is to make sure it doesn't grow at your expense.

The July 2025 launch of SAP Business AI in its current form, combined with SAP's 2025 AI and data licensing restructuring, set in motion a multi-year commercial shift that will be substantially complete by July 2026. Enterprises that understand the mechanics, build the governance, and negotiate the right contractual protections now will control their AI spend. Those that don't will fund SAP's AI revenue growth.

SAP's Pricing Model Is Changing — What That Means for Enterprise Budgets

The transition from named-user licensing to consumption-based AI pricing is not an incremental change in how SAP invoices its customers. It is a structural shift in the commercial relationship between SAP and its enterprise customers — one that fundamentally changes the budget risk profile, the governance requirements, and the negotiation dynamics of every SAP commercial relationship.

Under the named-user model, SAP's commercial upside was primarily driven by user count growth and maintenance escalation. Both were relatively predictable and manageable. Under the use-based AI model, SAP's commercial upside is driven by consumption — how intensively your users interact with AI features. This is inherently less predictable, because AI consumption scales with user behaviour, AI adoption rates, and the scope of AI-powered workflows, all of which are subject to organisational and operational dynamics that finance teams cannot easily forecast.

The implication for enterprise budgeting is significant. SAP is no longer a cost item where the annual budget can be set with reasonable confidence based on user count and contractual escalators. SAP AI costs require a consumption forecast that factors in deployment velocity, use case scope, and unit rate — and that forecast needs to be monitored monthly against actuals. For most enterprise finance teams, this is a new capability requirement that SAP's licensing model is imposing unilaterally.

The July 2025 AI and Data Licensing Changes

SAP's July 2025 announcement restructured the Business AI licensing framework in ways that are still being absorbed by enterprise buyers. The key changes were:

How Use-Based AI Pricing Actually Works

The operational mechanics of use-based AI pricing create several specific financial risks that enterprise buyers need to understand and mitigate.

Consumption-triggered invoicing. Unlike named-user licences — which generate a fixed monthly invoice — AI Unit overages generate additional invoices when your committed allocation is exhausted. These overage invoices can arrive outside the normal SAP billing cycle, creating budget management challenges for enterprises that allocate SAP spend annually and monitor it quarterly.

Invisible consumption drivers. Many AI Unit consumption events are triggered by background processes, not explicit user actions. An AI-powered workflow automation in S/4HANA may consume AI Units every time it processes a document, regardless of whether a user actively invoked the AI feature. Enterprises that deploy AI-powered automation at scale can generate significant AI Unit consumption from background processes that are invisible to the users and managers who authorised the deployment.

Adoption-driven cost acceleration. The organisational dynamics that drive AI adoption — user enthusiasm, change management programmes, process redesign initiatives — are exactly the dynamics that accelerate AI Unit consumption beyond forecast. Enterprises that succeed in driving AI adoption often find that their AI cost base grows faster than their AI value realisation, creating a unit economics challenge that SAP's pricing structure is designed to exploit.

The agentic AI multiplier. As SAP deploys Joule Agents for autonomous task execution, the per-task AI Unit consumption is significantly higher than for interactive Joule queries. An agentic workflow that processes 1,000 procurement approvals per day can consume more AI Units in a week than a Joule query deployment generates in a month. Enterprises that are planning agentic AI deployments need to build agentic consumption into their AI Unit forecasts now — before they sign agreements that don't account for this step change.

The Financial Impact: From Predictable to Variable

The financial impact of SAP's AI pricing shift is being felt unevenly across enterprise customers. Organisations with limited AI deployment are experiencing minimal impact — Base tier AI Units are sufficient, and consumption remains within committed allocations. But enterprises at the vanguard of SAP AI adoption — deploying Joule broadly, activating agentic workflows, and integrating AI across multiple SAP products — are discovering that their SAP AI cost base is growing at a rate that their original deal economics did not contemplate.

The pattern we observe in advisory engagements is consistent: an enterprise signs a RISE or S/4HANA Cloud renewal with a Business AI commitment that looks reasonable at deal signature. Within 12–18 months of AI deployment, consumption exceeds the committed allocation. The first overage invoice arrives. The procurement team escalates. And then begins an unplanned negotiation with SAP to restructure AI Unit pricing — from a position of operational dependence rather than commercial strength.

This is precisely the commercial scenario SAP's pricing model is engineered to create. Enterprises that understand this and prepare for it — by building consumption governance, negotiating protective contract terms, and maintaining credible alternatives — avoid this scenario. Those that don't, fund it.

The July 2026 Deadline: What Changes and Why It Matters

July 2026 is the commercial milestone by which SAP intends to have substantially migrated its cloud customer base to the new AI and data licensing framework. This has specific implications for enterprises with contract renewals scheduled in H2 2026 or early 2027: the terms SAP will offer at those renewals will be structured under the new framework, and the negotiating window to influence those terms is now.

The practical implications of the July 2026 migration include: AI features that are currently available on a fixed-fee basis under legacy contract structures will move to AI Unit consumption pricing; Data licensing components (SAP Datasphere, SAP Analytics Cloud AI features) will be explicitly metered and billed on a consumption basis; and enterprises that have not yet signed Business AI agreements will be offered them as part of their renewal package — with AI Unit commitments as the primary AI cost structure.

For enterprises currently on legacy SAP ERP (ECC) with traditional named-user licences and annual maintenance, the July 2026 deadline reinforces SAP's existing commercial pressure to transition to the cloud. The AI pricing model is only available in the cloud; on-premise customers cannot access AI features under the new framework without some form of cloud transition. SAP is using this as a migration accelerator — and it's working.

Governance and Budgeting Requirements

Managing use-based AI pricing requires governance capabilities that most enterprise procurement and ITAM functions have not yet built. The minimum governance framework for SAP AI cost control includes: a designated SAP AI licensing owner with responsibility for consumption monitoring and budget management; monthly consumption reports from SAP (or from SAP for Me) broken down by product, feature, and business unit; a budget allocation model that separates committed AI Unit costs from variable overage risk; and an internal approval process for AI deployments that includes consumption estimation as a mandatory step before deployment authorisation.

SAP's monitoring tools — primarily SAP for Me and the BTP cockpit — provide consumption data, but in forms that require significant interpretation and that don't automatically generate budget alerts. Building internal tooling or processes to operationalise consumption monitoring is a governance investment that enterprises need to make before — not after — AI deployment at scale. For more on building internal SAP governance, see our guide on building a SAP Licence Management Office.

What Enterprise Buyers Should Do Right Now

If your SAP contract is renewing before July 2027, the following actions are time-sensitive:

• Audit your current AI exposure: Review what AI features are active in your SAP environment, what your current AI Unit consumption is, and when your committed allocation will be exhausted at current burn rates.
• Map your renewal timeline: Identify when your SAP contract renewal occurs, when the notice deadline is, and how long before renewal you need to begin commercial negotiations.
• Build a consumption forecast: Model AI Unit consumption for your planned AI deployments over the next 3 years. This forecast should inform your renewal negotiation position — not SAP's.
• Engage independent advisory support: The July 2026 deadline creates a specific negotiation window. Independent advisors who understand SAP's AI pricing model can help you position for that window effectively. See our RISE with SAP advisory service and licence optimisation service for how we approach this.
• Evaluate your on-premise transition position: If you are on ECC, the AI pricing model is an additional data point in the cloud migration business case. Evaluate it as such — not as a reason to rush into a RISE transition without appropriate commercial protections.

Negotiation Priorities for the New Pricing Model

The transition to use-based AI pricing creates specific negotiation opportunities that do not exist in traditional named-user licence renewals. The commercial leverage is at its highest in the current window — before the July 2026 migration is complete and before SAP's renewal terms harden around the new model.

Key negotiation priorities include: consumption caps with hard limits on AI Unit overages (preventing unlimited financial exposure from consumption growth); price freezes on AI Unit rates for the duration of the contract term (SAP's default is to apply annual rate escalators to AI Units, compounding the cost growth from consumption increases); rollover rights for unused AI Units (the no-rollover default is commercially punitive and negotiable); and multi-year volume commitments in exchange for materially lower per-unit rates and the governance protections described above. For a complete view of SAP AI contract negotiation, including specific language to demand in your Order Form, see our companion guide. You can also review our broader contract negotiation service for how we structure these engagements.