How to prevent SAP Business Technology Platform costs from spiralling. From licensing model selection and credit forecasting to consumption governance, negotiation tactics, and the FinOps framework that saves enterprises 20-40% on BTP spend.
This guide is part of our SAP Advisory series. See also: BTP Licensing Models Explained | BTP Custom Development Costs | SAP Cloud and Hybrid Strategies.
SAP's Business Technology Platform (BTP) has become the strategic foundation for enterprise innovation on SAP, powering custom development, integration, analytics, AI/ML, and process automation. But BTP's consumption-based pricing introduces a fundamentally different cost dynamic. Unlike perpetual licences or fixed subscriptions, BTP costs are driven by actual resource consumption that can fluctuate dramatically, scale unpredictably, and generate budget overruns without active governance.
| BTP Cost Challenge | Typical Impact | Root Cause | Optimisation Opportunity |
|---|---|---|---|
| Over-committed credits expiring unused | 15-25% of annual commitment wasted | Optimistic forecasting; slow project adoption | Right-size commitment + credit acceleration programme |
| Uncontrolled consumption overage | 20-50% premium on list-price overages | No consumption alerts; unmonitored auto-scaling | Alert thresholds + auto-scale caps + monthly governance |
| Idle dev/test/sandbox environments | $50K-$300K+ annually in wasted credits | Always-on environments consuming 24/7 | Automated shutdown schedules + resource lifecycle policies |
| Wrong licensing model per workload | 10-20% cost premium vs optimal model | Default to one model without workload analysis | Per-workload model selection (CPEA vs subscription vs PAYG) |
| Weak enterprise agreement terms | No rollover, no overage protection, no flexibility | Insufficient negotiation preparation | Negotiated rollover, overage discounts, mid-term adjustment |
| Model | Cost Structure | Best For | Unit Cost | Flexibility | Risk |
|---|---|---|---|---|---|
| CPEA / BTP Enterprise Agreement | Prepaid credit pool; annual commitment ($150K+ minimum) | Multi-service, variable workloads at scale | Lowest (volume discounts) | High: credits across services | Unused credits expire; overage at premium |
| Subscription | Fixed fee per service; defined capacity | High-utilisation, predictable, single-service | Low (for high utilisation) | Low: locked to specific service | Pay for unused capacity; cannot flex down |
| Pay-As-You-Go | Per-use billing; no commitment | Low volume; experimental; unpredictable | Highest (no discounts) | Maximum: no lock-in | Cost can spike unexpectedly with high usage |
| Hybrid (CPEA + Subscription) | Credits for variable + subscriptions for stable | Mature BTP estates with mixed workloads | Optimised per workload | Balanced | Complexity in tracking two models |
SAP's standard CPEA terms expire unused credits at the end of each annual period. This means accurate forecasting is essential. Over-committing wastes money, under-committing triggers expensive overage rates. Different BTP services consume credits at dramatically different rates per unit of work. Mapping planned service usage to SAP's credit consumption tables is a prerequisite for accurate commitment sizing.
| Workload Type | Predictability | Volume | Recommended Model | Rationale |
|---|---|---|---|---|
| Production Integration Suite (CPI) | High | High | Subscription | Steady, high-volume; locked-in rate cheaper than credits |
| SAP HANA Cloud (production) | High | High | Subscription | Always-on production database; predictable capacity |
| SAP Build Apps (citizen dev) | Low | Variable | CPEA credits | Unpredictable adoption; usage spikes during sprints |
| SAP AI Core / AI Launchpad | Low | Variable | CPEA credits | Experimental; consumption depends on model training cycles |
| SAP Analytics Cloud (production) | Medium | Medium | CPEA or Subscription | Depends on user count stability; model both options |
| Development/sandbox environments | Low | Low-Medium | CPEA credits | Variable usage; should be shut down when idle |
| Initial BTP exploration/POC | Very low | Low | Pay-As-You-Go or free tier | No commitment until usage patterns established |
Subscribe to high-volume, predictable production services (Integration Suite, HANA Cloud) at locked-in rates, and use CPEA credits for everything else: development environments, new services, variable workloads, and experimentation. This captures subscription efficiency for known workloads while maintaining credit flexibility for everything less predictable. Model both approaches for each workload with actual consumption data before committing.
| Alert Level | Threshold | Action Required | Owner |
|---|---|---|---|
| Green | 0-50% consumed on schedule | Informational; no action needed | BTP Admin (monthly review) |
| Amber | 75% consumed; ahead of plan | Review remaining projects; identify savings opportunities | FinOps Team (within 1 week) |
| Red | 90% consumed; significantly ahead | Escalate to leadership; decide on additional credits vs curtailment | CIO / CFO (within 48 hours) |
| Critical | 100% consumed; overage active | Immediate action: shut down non-essential; engage SAP for additional credits | CIO / Procurement (immediate) |
Accurate credit forecasting requires three inputs: historical consumption data (minimum 3 months, ideally 12 months), planned project pipeline (new BTP workloads coming online), and seasonal patterns (fiscal close, year-end processing). Project credit burn rate against the remaining pool. Update monthly with actual data. This is the single most effective tool for preventing both wastage and overage.
| Technique | How It Works | Typical Savings | Effort |
|---|---|---|---|
| 1. Shut down idle dev/test environments | Development, test, sandbox, and demo environments running 24/7 consume credits continuously. A dev HANA Cloud instance running overnight and weekends consumes 128 hours/week when only 40-50 hours of active use occur. Implement automated shutdown schedules | 15-25% of total BTP spend | Low-Medium (2-4 weeks) |
| 2. Right-size service instances | Over-provisioning (selecting larger instance than needed) wastes credits on unused capacity. Review actual utilisation for every service instance. Downsize where utilisation is consistently below 50% | 15-30% per over-provisioned service | Medium (2-6 weeks) |
| 3. Control auto-scaling | BTP services with auto-scaling (Integration Suite, Kyma runtime, HANA Cloud compute) can scale up automatically, consuming credits at higher rates. Set explicit maximum scaling limits for every auto-scaling service | Prevents 10-30% cost overruns during peaks | Low (1-2 weeks) |
| 4. Leverage free tier and trial plans | SAP offers free tier allocations and trial plans for many BTP services. Use for prototyping, POC development, training, and non-production experimentation before consuming paid credits | $20K-$100K+ annually | Low (immediate) |
| 5. Consolidate subaccounts and eliminate redundancy | Multiple subaccounts created by different teams each provisioning their own service instances creates duplicated resources. Audit all subaccounts; consolidate where architecturally feasible | 10-20% through duplication elimination | Medium-High (1-3 months) |
| 6. Optimise Integration Suite message processing | CPI is often the highest-consumption service. Reduce unnecessary API calls, implement caching, batch messages, eliminate polling in favour of event-driven integration | 15-25% of Integration Suite consumption | Medium (2-8 weeks) |
| 7. Implement resource lifecycle management | Resources created for projects or POCs persist after project ends, consuming credits. Implement mandatory resource tagging (project, owner, expiry) and automated cleanup policies | 5-15% of total consumption | Medium (1-2 months) |
| Negotiation Point | SAP Default | Target Enterprise Position | Financial Impact |
|---|---|---|---|
| Volume discount | Published tier rates | 15-30% additional discount for $1M+ commitment | $150K-$300K savings per $1M committed |
| Overage rates | On-demand rates (20-50% premium) | Committed rate + 5-10% for first 20% overage | Prevents $50K-$200K in overage premiums |
| Credit rollover | No rollover; credits expire | 10-20% rollover to next period | Recovers $50K-$200K+ in expiring credits |
| Mid-term adjustment | Fixed commitment for full term | Plus/minus 15-20% adjustment at Year 1 checkpoint | Prevents over-commitment waste or under-commitment overage |
| Service portability | Generally broad; verify specifics | Explicit language covering all current and future BTP services | Ensures maximum credit flexibility |
| Strategy | When to Use | Detail |
|---|---|---|
| 90-day credit review | 90 days before annual expiry | Formal review comparing remaining credits against projected consumption. If surplus projected, activate credit acceleration. If deficit, plan additional purchase at negotiated rates |
| Credit acceleration programme | When surplus identified | Pull forward planned projects and workloads to consume credits productively: scheduled migrations to Integration Suite, HANA Cloud data loading, developer training, POC/pilot projects |
| Strategic one-time workloads | When acceleration insufficient | Deploy productive one-time workloads: advanced analytics/AI training, data quality assessments, performance load testing, reusable integration templates. Any productive use is better than expiry |
| Negotiate rollover | If rollover provisions exist in contract | Document unused amount and formally request rollover. If no provisions exist, use unused credits as leverage at next renewal to negotiate smaller, better-fitted commitment |
Target 90-95% utilisation. Below 85% indicates systematic over-commitment that should drive a smaller commitment at the next renewal. Set the 90-day review in the calendar today for every BTP credit period. This is the minimum lead time needed to activate acceleration or purchase additional credits.
| Governance Cadence | Activity | Participants | Key Output |
|---|---|---|---|
| Daily | Automated consumption monitoring; alert processing | BTP Admin | Alerts triggered if thresholds breached |
| Monthly | Cost review: actual vs forecast; anomaly analysis; optimisation status | FinOps + BTP Team + Business | Monthly cost report; action items |
| Quarterly | Credit utilisation review; forecast update; model optimisation | FinOps + IT Leadership | Updated forecast; model adjustment recommendations |
| 90 days before expiry | Surplus/deficit review; credit acceleration or additional purchase | Finance + Procurement + IT | Credit utilisation plan for remaining period |
| 60-90 days before renewal | Annual commitment planning; negotiation preparation | Procurement + FinOps + Advisory | Recommended commitment size; negotiation strategy |
Real-time visibility (dashboards across all subaccounts showing consumption vs remaining pool, burn rate vs forecast). Accountability and cost allocation (assign BTP costs to teams/projects, chargeback/showback reporting). Monthly governance review (actual vs forecast, top 5 cost drivers, upcoming project impact). Annual commitment planning (60-90 days before renewal, based on 12 months actual data, target 90-95% utilisation).
| Strategic Consideration | Detail |
|---|---|
| BTP within RISE | RISE bundles typically include base BTP credits. Understand exactly how many are included and whether they cover your needs. If insufficient, negotiate additional credits within the RISE agreement rather than separately. Bundled pricing is consistently better than standalone BTP agreements |
| Integration cost between BTP and S/4HANA | BTP is SAP's designated integration and extension platform for S/4HANA. Integration scenarios that were handled within ECC now require BTP services with consumption costs. When planning S/4HANA migration, model BTP integration costs explicitly. Frequently underestimated, can represent 15-25% of total S/4HANA Cloud cost |
| Total SAP cloud cost governance | Span BTP credits, RISE/S/4HANA subscriptions, SaaS products (SuccessFactors, Ariba, Concur), and SAP infrastructure costs. Holistic view prevents optimisation in one area from increasing costs in another. Also maximises negotiation leverage by presenting SAP with consolidated commercial relationship |
If you are migrating to S/4HANA (Cloud or on-premises), BTP is SAP's designated platform for extensions, integrations, and custom development. BTP costs should be modelled as part of the S/4HANA business case, not as a separate, unexpected add-on. Negotiate BTP as part of your largest SAP deal for maximum leverage.
| # | Action | Timeline | Expected Impact |
|---|---|---|---|
| 1 | Inventory all BTP services: list every service instance across all subaccounts with tier, capacity, and consumption data | Week 1-2 | Foundation for all optimisation actions |
| 2 | Map workloads to optimal licensing model: assess each against CPEA/subscription/PAYG decision framework | Week 2-4 | 10-20% savings through model optimisation |
| 3 | Implement automated shutdown schedules for all non-production environments | Week 2-4 | 15-25% reduction in non-production consumption |
| 4 | Right-size all over-provisioned service instances: reduce tier/capacity where utilisation below 50% | Week 3-6 | 15-30% per-service savings |
| 5 | Set auto-scaling maximums; configure consumption alerts at 50%, 75%, 90%, 100% thresholds | Week 2-3 | Prevents 10-30% cost overruns |
| 6 | Consolidate redundant subaccounts and eliminate duplicated service instances | Month 2-4 | 10-20% through duplication elimination |
| 7 | Implement resource tagging and lifecycle management: mandatory tags, automated orphan detection | Month 2-3 | 5-15% through orphaned resource cleanup |
| 8 | Deploy real-time consumption dashboards with cost attribution by subaccount, project, and team | Month 2-4 | Enables accountability-driven optimisation |
| 9 | Establish monthly BTP cost governance review: actual vs forecast, anomaly analysis, optimisation status | Ongoing monthly | Sustains all savings; prevents cost creep |
| 10 | Prepare renewal negotiation: 12 months consumption data, optimal commitment model, targets for discounts/rollover/overage protection | 60-90 days before renewal | 15-30% better renewal terms |
No single best model. CPEA (cloud credits) is best for multi-service, variable workloads at scale. Subscriptions are best for high-volume, predictable, single-service workloads. Pay-As-You-Go for low-volume or experimental. Most enterprises benefit from a hybrid: subscriptions for known production workloads, CPEA credits for everything else.
Not by default. SAP's standard CPEA terms expire unused credits at each annual period end. However, rollover of 10-20% is negotiable for large enterprise agreements, particularly when bundled with multi-year commitments or RISE agreements. Even without rollover, unused credits should drive a smaller commitment at the next renewal.
Four layers of protection: consumption monitoring with threshold alerts (50%, 75%, 90%, 100%), automated shutdown of non-production environments outside business hours, auto-scaling maximum limits on all scalable services, and monthly governance reviews comparing actual consumption against forecast. Enterprises implementing all four consistently report 20-40% lower BTP costs.
Base on 12 months of actual consumption data (or 3+ months if newer), plus projected consumption from planned new workloads, adjusted for seasonal patterns. Target 90-95% utilisation: high enough to maximise volume discounts but with buffer to avoid overage. Negotiate a mid-term adjustment clause to right-size if actual diverges from forecast.
Five most common: idle dev/test environments running 24/7 (15-25% of total spend), over-provisioned service instances (15-30% per service), uncontrolled auto-scaling (10-30% overruns), orphaned resources from completed projects (5-15%), and over-committed credits expiring unused (15-25%). All five are addressable through the techniques and governance framework in this guide.
RISE agreements typically include base BTP credits. Understand exactly how many are included, whether they cover your needs, and negotiate additional credits within the RISE agreement rather than separately. BTP integration costs for S/4HANA should be modelled explicitly. They are frequently underestimated and can represent 15-25% of total S/4HANA Cloud cost.
Highest consumption typically: SAP Integration Suite (message processing volume), SAP HANA Cloud (memory and compute), SAP AI Core (model training and inference), SAP Build Apps (runtime and storage), and SAP Analytics Cloud (query processing and user access). Prioritise optimisation efforts on your highest-consumption services for maximum impact.
For BTP commitments exceeding $500K annually, independent advisory typically delivers 5-10x ROI. SAP's cloud sales teams are expert negotiators. Independent advisors bring current market benchmarking, negotiation tactics specific to CPEA/BTP commercial terms, and the ability to identify clauses (rollover, overage protection, mid-term adjustment) that internal teams may not know are achievable.
Redress Compliance provides independent advisory on SAP BTP licensing, consumption optimisation, and enterprise agreement negotiation. Deep expertise in CPEA commercial models and BTP FinOps governance. 100% vendor-independent. Fixed-fee engagement.
SAP Advisory ServicesIndependent SAP advisory. BTP consumption optimisation. CPEA negotiation. FinOps governance. 100% vendor-independent, fixed-fee engagement.