Two Cost Models: CapEx vs OpEx
The SAP on-prem vs RISE decision is a choice between two fundamentally different cost shapes. On-premise S/4HANA is a capital purchase: you buy a perpetual licence once, then pay SAP Enterprise Support at 22% of net licence value every year, and you carry the infrastructure yourself. RISE with SAP is an operating expense: a per-FUE subscription that bundles the S/4HANA software, the hyperscaler infrastructure, and support into a single recurring fee, with SAP running the technical operations.
That difference is not just an accounting preference. Perpetual licensing leaves you owning an asset whose marginal cost falls over time, but exposes you to hardware refresh cycles, a Basis team, database administration, backup and disaster recovery, and security monitoring. RISE removes those operational line items but replaces a depreciating asset with a subscription that escalates — typically 5–7% per year at renewal. Before modelling either, it is worth grounding the licensing mechanics in our SAP licensing guide, and deciding the edition question first in our RISE vs GROW comparison.
The 4–7 Year TCO Crossover
RISE looks cheaper early because it defers the big numbers. With no upfront perpetual licence and no hardware to buy, the first three to five years of a RISE subscription often land below the equivalent on-premise spend — and SAP markets some cloud packages at roughly 20% lower five-year TCO precisely because the comparison window is set short. The picture changes once the perpetual licence has been amortised: beyond the crossover, on-premise marginal cost is essentially the 22% maintenance plus infrastructure, while the subscription keeps escalating.
Across our engagements and published benchmarks, the TCO crossover typically falls between years four and seven. The exact point depends on three variables: the perpetual discount you negotiated, how efficiently you can staff and refresh infrastructure, and the subscription escalator SAP applies. The shorter your planning horizon, the more RISE flatters itself; the longer you intend to run the system, the more a well-bought perpetual licence pulls ahead.
SAP's standard RISE business case compares the subscription against your current on-premise run-rate over a five-year window — the one window in which the cloud almost always wins. Re-run the same model over seven and ten years before you accept the conclusion.
A Worked 500-User Example
Consider a 500-user S/4HANA estate. On-premise, a perpetual licence of roughly $5M plus $1.1M per year in Enterprise Support totals about $10.5M over five years before infrastructure. RISE at a list rate near $150 per user per month works out to $4.5M over five years, falling to about $2.7M with a 40% negotiated discount — a clear five-year win for the cloud.
| Cost element | On-Premise (perpetual) | RISE (subscription) |
|---|---|---|
| Upfront licence | ~$5M one-time CapEx | None |
| Annual run cost | $1.1M support (22%) + infra | ~$2.5M subscription, +3–7% / yr |
| 5-year TCO | ~$10.5M+ (pre-infra) | ~$2.7M (40% discount) |
| 10-year TCO (modelled) | ~$16M | ~$28.7M |
| Hidden ops costs | Carried by you (Basis, DR, HW) | Bundled in subscription |
| Asset at end of term | Owned perpetual licence | Nothing — rights lapse |
Extend the same model to ten years and it inverts: on-premise lands near $16M (licence plus a decade of maintenance and infrastructure) while the cloud equivalent at $2.5M per year with 3% escalation reaches roughly $28.7M. The caveat matters — on-premise TCO can run 20–40% higher than private cloud once you fully load the hidden operational costs, so the gap narrows for organisations that cannot staff or refresh infrastructure efficiently. The point is not that one always wins; it is that the answer is entirely sensitive to the horizon and the discount, both of which you control.
Conversion Credits and the 2027 Clock
If you are moving an existing ECC or S/4HANA perpetual estate, conversion credits change the maths. SAP applies a conversion credit of typically 50–70% of the perpetual residual value against the RISE subscription in years one to three, falling to zero by year four — and the underlying residual depreciates at roughly 10% per year. The credit is real money, and it is largest when you convert deliberately rather than under pressure.
The pressure is the 31 December 2027 end of ECC mainstream support, the single strongest piece of leverage in any S/4HANA negotiation — and a double-edged one. Converting under deadline duress, with an expiring contract forcing your hand, hands the timing advantage back to SAP. Converting on your own clock, with the credits modelled and a target close around SAP's Q3 (July–September) when discount authority peaks, captures the offset in full. Our S/4HANA migration negotiation and RISE contract negotiation guides set out how to sequence that, and the SAP vendor hub tracks the credit and pricing changes worth watching.
Which Model Wins — and How to Test It
Favour RISE if your planning horizon is five years or less, you cannot efficiently staff Basis and infrastructure, you value predictable OpEx and offloaded operations, and you can negotiate a deep enough discount and conversion credit to keep the five-to-seven-year curve competitive. Favour on-premise perpetual if you intend to run the system well beyond seven years, you already operate infrastructure at scale, and you can secure a steep perpetual discount — because owning a paid-down asset with a 22% marginal cost is hard to beat over a long horizon.
Whichever way you lean, do not accept SAP's business case as built. Re-run it over seven and ten years, load every hidden operational cost on the on-premise side and every escalator on the cloud side, and model the conversion credit at its actual depreciation. The model that survives that scrutiny is your real answer. To pressure-test your own numbers against live deal data, request a confidential briefing and download the SAP S/4HANA Guide.