The S/4HANA migration decision facing enterprise finance and IT teams is rarely about capability. SAP's newest ERP platform is undeniably superior. The real question is affordability—and it's one that vendor marketing budgets actively obscure.
In my role defending enterprise clients in licensing disputes over indirect access, license metrics, and audit findings, I've seen the financial consequences of poor TCO analysis. The cost structures here are genuinely complex, with variables that compound across five-year windows. A mid-market manufacturer I advised last year underestimated their total migration spend by $1.2 million. Their initial budget: $250,000. Their actual cost: $1.45 million.
This analysis compares the genuine financial profiles of both paths—on-premise S/4HANA Professional licensing with perpetual rights, and RISE with SAP's managed cloud offering. I'll show you what each approach actually costs, where the hidden line items hide, and the framework CIOs use to make this decision correctly.
The Real Migration Equation
SAP's migration economics have shifted dramatically since 2024. The vendor still offers migration credits to ease the transition from ECC and older S/4HANA versions. But here's what SAP doesn't emphasize: those credits decrease approximately 10 percent annually. A company migrating in 2025 receives significantly more credit than one migrating in 2027.
Migration costs themselves are wildly variable. Horváth & Partners surveyed 200 large companies implementing S/4HANA. The results were sobering: only 37 companies completed their migrations on time and within budget. More than 60 percent reported exceeding their budgets and slipping their timelines. The complexity multiplier isn't exotic technology—it's data structure conversion, legacy system decommissioning, user change management, and parallel-run infrastructure that vendors routinely underestimate.
The Horváth data shows that complexity-light migrations cost $250,000 to $1.5 million. Mid-complexity migrations run $1.5 million to $4 million. Full-scale transformations with deep customization refactoring exceed $8 million. Industry consultants will tell you similar numbers, and their incentive is actually to go conservative—they get blamed if actuals exceed estimates.
A critical hidden cost that no vendor marketing material mentions: dual-run periods. You typically run ECC and S/4HANA in parallel for 3 to 12 months, depending on your industry and regulatory requirements. During that window, you're paying for infrastructure, personnel time, and system maintenance across both platforms. This effectively doubles infrastructure costs during transition.
If your baseline infrastructure cost is $150,000 per month, a 9-month dual-run period adds $1.35 million in incremental spending. Few companies budget for that explicitly, but every finance team experiences it.
On-Premise S/4HANA: What You Actually Pay
On-premise licensing breaks into several components. Professional Users cost $3,000 to $6,000 per user as a one-time license fee. That gives you perpetual rights to that version. It does not include future major releases—you pay again for those upgrades—but it does establish a fixed asset that never expires.
Annual maintenance on that license runs 22 percent of license value per year. So a $5,000 per-user license costs $1,100 annually in support and patches. That percentage continues indefinitely. After about five years, cumulative maintenance roughly equals your initial license cost.
Infrastructure costs depend heavily on your architecture. Organizations with strong internal IT typically see hardware and cloud IaaS costs of $80,000 to $300,000 per year for mid-market deployments (1,000 to 2,000 concurrent users). Enterprise deployments scale higher, obviously.
Database licensing is a separate line item. S/4HANA typically runs on either SAP HANA (SAP's proprietary database) or third-party databases like SQL Server. HANA licensing costs roughly $150,000 to $400,000 per year depending on data volume and processor consumption. SQL Server typically costs less—$30,000 to $100,000 annually—but carries additional compliance considerations in some industries.
Implementation and go-live support add $500,000 to $2 million depending on scope. Partner services during years one and two are almost universally necessary.
A five-year on-premise S/4HANA TCO for a 1,500-user mid-market company looks roughly like this:
- License acquisition (1,500 users at $4,000 avg): $6,000,000
- Database licensing (five years): $900,000
- Annual maintenance (22% of license value): $1,100,000
- Infrastructure (cloud IaaS or on-premise): $600,000
- Migration and implementation: $1,200,000
- Internal IT staffing (dedicated resources): $800,000
- Five-year total: $10,600,000
That spreads to roughly $2.12 million per year. On a per-user basis for 1,500 users, you're at approximately $1,413 per user per year across the entire cycle—but heavily front-loaded.
Important nuance: if you're already running ECC, you already have infrastructure and IT staffing. The incremental cost of S/4HANA migration in that scenario is lower—you're primarily adding license costs and migration work, not starting from zero infrastructure.
RISE with SAP: What's In and What's Not
RISE with SAP bundles S/4HANA Private Cloud, Business Technology Platform (BTP) credits, and infrastructure management into a single monthly subscription. For mid-market deployments (1,000 to 2,000 users), pricing runs $147 to $178 per Fully Utilized Employee (FUE) per month. That's effectively a per-user metering model disguised as a platform subscription.
The bundle includes the cloud system itself, standard maintenance and support, infrastructure management, and a baseline allocation of BTP credits for workflow extensions and analytics. The pricing is typically fixed for a three-year contract period, with documented price increase caps—usually 3 percent annually, though SAP has flexibility in how those caps work across different customer cohorts.
Here's what RISE does not include: custom development beyond BTP's platform-as-a-service model. If your business logic lives in custom Z-tables, custom function modules, or complex ABAP modifications, RISE requires you to refactor those into BTP applications. That refactoring work is expensive—often $100,000 to $1 million depending on complexity. SAP calls this the "Clean Core" strategy. It's genuinely sensible architecture, but it's also a cost that most vendors don't highlight in marketing materials.
Additional BTP credits beyond the bundled allocation cost roughly $0.30 to $0.60 per credit-hour. For companies with heavy analytics workloads or significant custom application logic, that adds up quickly.
The bigger transparency issue: RISE subscription prices are not locked to perpetuity. SAP can increase prices in future contract periods. Unlike perpetual on-premise licensing—where you own the rights forever—RISE effectively converts your software spend to perpetual renting. Five years from now, your per-user cost could be 20 percent higher.
Five-year RISE TCO for the same 1,500-user mid-market company:
- RISE subscription (1,500 FUE at $160/month average): $14,400,000
- Beyond-bundled BTP credits: $300,000
- Custom application refactoring (Clean Core): $600,000
- Internal IT staffing (reduced scope): $400,000
- Migration support (included in subscription year 1): $0
- Five-year total: $15,700,000
That's approximately $3.14 million per year, or $2,093 per user annually.
SAP commercial advisory specialists can help model your specific scenario with real pricing
Get a personalized TCO assessment for your user count and complexity profileFive-Year TCO Comparison
On-premise costs $10.6 million. RISE costs $15.7 million. But those numbers hide important nuances that swing decisions.
On-premise TCO is front-loaded. Years one and two consume 45 percent of the five-year budget due to migration work and implementation. Years three through five are significantly cheaper—just running and maintaining an owned system. If your company's financial horizon is uncertain, the later-period cost certainty is genuinely valuable.
RISE TCO is linear. Year one through five costs remain roughly consistent (aside from contracted price increases). For companies with volatile cash flow or uncertain user growth, that predictability has real value. Finance teams love stable, predictable operating expenses.
But here's the financial inflection point: at five years, RISE still owns zero intellectual property in the system. Your customizations live in BTP, which SAP owns. Your data lives on SAP's infrastructure. If you walk away after five years, you walk away with nothing proprietary—a clean separation, but also zero asset value.
On-premise? After five years, you own a perpetual license. You've paid $10.6 million, and you have rights to run that system indefinitely. Year six costs drop dramatically—just annual maintenance at 22 percent of license value, roughly $1.32 million. That $10.6 million investment begins paying dividends in years six through ten. Over a ten-year window, on-premise total cost is approximately $13.9 million. RISE over ten years, assuming 3 percent annual price increases, reaches approximately $18.5 million.
The crossover point depends on your actual pricing, but most models show on-premise becoming cheaper around year seven or eight, and substantially cheaper by year ten.
Hidden Costs SAP Won't Discuss
Several line items almost never appear in vendor TCO models.
ECC maintenance extension. SAP mainstream support for ECC ends December 31, 2027. But extended maintenance runs through 2030. Extended maintenance costs approximately 24 percent of license value annually—a 2 percent uplift versus standard 22 percent maintenance. If you're on ECC and thinking you can delay migration, that extended maintenance cost is real. Horváth research shows 60 percent of large companies won't complete migration before that 2027 deadline.
Third-party support economics. Companies like Rimini Street provide maintenance support for SAP systems at 40 to 50 percent below SAP's published rates. If you migrate to on-premise S/4HANA and then switch to Rimini Street for year two onward, you can reduce maintenance costs significantly. That's not a hidden cost—it's actually cost elimination—but most companies don't budget for it because SAP certainly doesn't mention it.
The new ERP Private Edition transition option. In Q1 2025, SAP introduced ERP Private Edition, a transitional licensing path for select large enterprises. It essentially extends ECC support through 2033 with new licensing economics. If you have more than 10,000 users, this might be a legitimate third option between on-premise S/4HANA and RISE. Most mid-market companies don't qualify, but you should check your eligibility.
Dual-run infrastructure (again, quantified). A 9-month parallel run on typical mid-market infrastructure adds $1.35 million in cost. That's the single largest "surprised us" line item I've seen in actual company budgets. Many finance teams assume parallel-run infrastructure is already budgeted for in implementation costs. It rarely is.
Migration credits decay. SAP migration credits are front-loaded. If you delay migration by two years, you potentially leave 20 percent of the credit on the table. That's another $300,000 to $1 million depending on your scenario. The financial incentive to migrate sooner is real.
Enterprise IT strategy requires accounting for all variables
Schedule a compliance consultation to review your licensing positionThe ECC Private Edition Option
SAP introduced a third path that's meaningful for select large enterprises: ERP Private Edition. It's designed for organizations with more than 10,000 users who genuinely can't migrate to S/4HANA by 2030, or who need extended deployment flexibility.
ERP Private Edition extends ECC support through 2033, with new licensing mechanics that essentially combine elements of traditional perpetual licensing with a managed services layer. Pricing is custom-negotiated, but typically runs lower than S/4HANA on-premise for large cohorts.
The trade-off: you're staying on an older technology platform. ECC itself has no material new functionality; you're essentially buying time. For companies with complex customization, aging infrastructure, or financial constraints, that can be rational. But it's explicitly a delay strategy, not a long-term platform choice.
If your company is considering ERP Private Edition, the TCO math changes. You might avoid significant migration costs but accept higher maintenance costs in years 2027 to 2033. The decision tree here is highly organization-specific.
Decision Framework for CIOs
Here's the actual decision framework that enterprise IT leaders should use:
Choose on-premise S/4HANA if:
- Your company has a 10+ year IT platform horizon and strong in-house database and infrastructure expertise
- You have extensive custom ABAP code that refactoring to BTP would make cost-prohibitive
- You operate in highly regulated industries (financial services, pharmaceuticals) with specific data residency or infrastructure requirements
- You have stable user counts and can front-load the migration cost in year one
- Your company has internal IT staffing with SAP expertise and wants to retain hiring/operational control
Choose RISE with SAP if:
- Your company wants to eliminate infrastructure management and delegate that to a vendor
- You expect significant user growth and prefer metering to upfront per-user licensing
- Your customizations are light or easily adaptable to BTP's platform model
- Your financial model prioritizes operational expense certainty over capital efficiency
- You have limited internal SAP expertise and prefer managed services
- You want SaaS-style rapid deployment and don't need the flexibility of on-premise
Consider ERP Private Edition if:
- Your company has more than 10,000 users and cannot migrate before 2030
- Your ECC customization is so extensive that S/4HANA migration is genuinely unaffordable in your planning window
- You need a transition bridge that delays major capital investment
Seven Questions Before You Sign
Before any executive sponsors either contract, ask these questions and demand clear answers:
1. What is the actual per-user cost over five years, including all services? Don't accept vendor marketing numbers. Build your own model with real implementation scope, infrastructure costs, and support levels. Include migration credits explicitly, with dates tied to actual timelines.
2. What happens to pricing in years 4 and 5 of a three-year RISE contract? If you're considering RISE, insist on understanding the price increase logic. SAP typically uses "published list price plus 3 percent per year" but that's negotiable. Push for longer-term caps if possible.
3. What is our actual total migration cost, with specific line items for dual-run infrastructure? This is where 30 percent of budget overruns live. Get implementers to specifically budget parallel-run infrastructure separately from go-live costs.
4. If we choose on-premise, what is our actual maintenance vendor strategy? Can you commit to SAP support, or do you want to evaluate third-party providers like Rimini Street after year one? That decision affects your five-year cost by 15 to 30 percent.
5. What is the depreciation and asset life accounting impact of each choice? On-premise capitalization allows depreciation; RISE is pure opex. This affects your balance sheet and P&L structure. Talk to your finance team about which is preferable given your organization's cash flow and accounting policies.
6. Do we qualify for any migration credits, and what is the timeline sensitivity? If credits expire by 2027, that creates urgency. Understanding the specific credit amount and decay schedule is financially material.
7. What is our long-term customization strategy? If you choose RISE, you're accepting BTP as your customization platform permanently. If you choose on-premise, you're accepting the cost of maintaining custom ABAP code. Which burden fits your IT strategy?
The S/4HANA decision is fundamentally a financial one wrapped in technology language. Get the numbers right, account for the hidden costs, and make the choice based on your specific five-year and ten-year financial profile, not on vendor marketing momentum.
I've seen hundreds of companies regret this decision after contract signature. Most regret it because they didn't account for dual-run infrastructure, didn't understand their actual customization refactoring costs, or didn't model the price increase trajectory on RISE over a longer time horizon. Don't be one of them. Build the model carefully.