The deployment model moved more money than any lever inside it, and exit economics appeared in fewer than half of first proposals
Four commercial models carry S/4HANA, and the decision between them is made once, early, and usually by default. Everything negotiated afterwards operates inside a frame that decision already set.
Prepared by Redress Compliance · August 16, 2026 · SAP advisory. 25 to 35 S/4HANA migration and licensing engagements, 2024 to 2025.
Executive summary
The model choice outweighs every lever inside it. On premise perpetual, RISE private cloud, GROW public cloud, and hyperscaler hosted perpetual carry different rights, different exit positions, and different five year costs.
SAP opened with RISE in nearly every cycle observed, regardless of the customer infrastructure strategy, which makes the default a vendor preference rather than a fit assessment.
Conversion credits varied by double digit percentages between comparable deals. It is a negotiated term rather than a published rate, and it is the largest single number in most conversions.
Exit economics were modelled in fewer than half of first proposals. Subscription models end usage rights at term end. Perpetual rights survive a dispute. That asymmetry frames the whole negotiation and is usually priced last or not at all.
Four models, and what each one actually grants
The deployment decision is a rights decision before it is a cost decision, and the rights differ more than the price sheets suggest.
| Model | What you hold | At term end | Where the cost concentrates |
|---|---|---|---|
| On premise perpetual | Owned licences plus support | Rights survive; support is optional | Infrastructure and internal operations |
| RISE private cloud | Subscription to a bundled managed service | Usage rights end | FUE pool, escalation, and the bundle |
| GROW public cloud | Subscription to a standardised service | Usage rights end | Standardisation limits and extension routes |
| Hyperscaler hosted perpetual | Owned licences on rented infrastructure | Rights survive; hosting is separable | Infrastructure contract, negotiated separately |
Subscription models end usage rights at term end. Perpetual rights survive a dispute. That single asymmetry frames every negotiation that follows, because it decides what happens if the relationship goes wrong. A buyer holding perpetual licences who disagrees with SAP has a running system and an argument. A buyer on subscription who disagrees with SAP has a deadline. Neither position is automatically correct, and only one of them is usually priced.
The default is a vendor preference, not a fit assessment
Across the S/4HANA engagements advised, the deployment model choice moved more money than any negotiation lever inside it, and it was routinely made without being treated as a decision. SAP account teams opened with RISE in nearly every cycle observed, regardless of the customer's infrastructure strategy, existing hyperscaler commitments, or operating model. That is not improper. RISE is SAP's strategic vehicle and presenting it first is what an account team is there to do. It becomes expensive when the buyer treats the opening position as the shortlist.
Two consequences follow from accepting the default early. The first is that the negotiation happens entirely inside a frame that has already been set, so the levers available are FUE sizing, escalation caps, and bundle scope, all of which matter but none of which change the shape of the commitment. The second is that the alternative stops being credible. Conversion credits varied by double digit percentages between comparable deals in our file, and the variance tracked how live the buyer's alternative was rather than anything about the estate. SAP negotiates hardest against the already decided, which is why keeping a second model genuinely alive until signature is worth more than any clause.
The exit question is where the asymmetry does its work, and it was modelled in fewer than half of first proposals we reviewed. On a subscription, usage rights end when the term does, so the cost of disagreement includes the cost of migrating an ERP estate under time pressure. On perpetual, the rights survive and the argument is about support rather than about continuity. Pricing that difference is not a theoretical exercise. It is the number that tells you what the renewal conversation will feel like in year four, and it belongs in the first model rather than the last.
Two more disciplines are worth naming because they are cheap and routinely skipped. Size FUE from measured authorisation usage rather than from contracted named users, since the contracted figure describes what was bought and the measured one describes what is used. And resolve indirect access exposure before it prices into a RISE envelope, because once it is inside the bundle it is being valued at SAP's number rather than argued on its merits. The RISE mechanics sit in the RISE pillar, the public cloud comparison in RISE against GROW, and the wider library in the SAP practice.
- 520 vendor benchmarks, from SAP RISE to Oracle ULA to Microsoft EA
- Exit economics modelled per model, including a time pressured migration
- Every risky clause flagged with the exact quote, the page, and the replacement language
What to settle before the model is chosen
- Model the exit at term end for each option, including what migrating an ERP estate under time pressure would actually cost, because that is what a subscription term is really pricing.
- Treat the conversion credit as negotiable, since it varied by double digit percentages between comparable deals and it is usually the largest single number in the conversion.
- Size FUE from measured authorisation usage, not from contracted named users, because the contract records what was bought rather than what is used.
- Resolve indirect access before it enters a RISE envelope, since inside the bundle it is priced at SAP's number rather than argued separately.
- Keep a credible second model alive until signature. SAP negotiates hardest against the already decided, and the conversion credit variance tracked exactly this.
- Separate the infrastructure contract where the model allows it, because hyperscaler hosted perpetual keeps that negotiation yours rather than bundling it at SAP rates.
What the S/4HANA decisions showed, 2024 to 2025
Across roughly 25 to 35 SAP S/4HANA migration and licensing engagements, the deployment model choice moved more money than any negotiation lever inside it:
First proposals that priced exit economics at term end, despite subscription models ending usage rights entirely.
Percentage spread in how existing perpetual licence value was credited across comparable deals.
RISE was presented as the default regardless of the customer's infrastructure strategy in nearly every cycle observed. That is the account team doing its job, and it becomes costly only when the opening position is mistaken for the shortlist.
The four models are not interchangeable and the differences are structural rather than commercial. Read the rights before the rates, because the rates are negotiable inside a model and the rights are not.
Watch the briefing · 4:11S/4HANA Negotiations: The Discount Is Dead. The Tiers Are Not.Where the money actually sits in an S/4HANA deal once the deployment model is fixed.
Your first five moves
- Price all four models before engaging, so the opening position is compared against alternatives rather than accepted as the frame.
- Model exit economics at term end for each, including a time pressured ERP migration, and put the number in the first business case.
- Rebuild the FUE count from measured authorisation usage rather than from contracted named users.
- Settle indirect access exposure separately before it prices into a bundled envelope at SAP's valuation.
- Keep the second model live until signature, because the conversion credit tracked how credible the alternative was. The SAP practice prices the models with you.
Frequently asked questions
What are the four S/4HANA commercial models?
On premise perpetual, RISE private cloud, GROW public cloud, and hyperscaler hosted perpetual. They differ in what you hold, what happens at term end, and where the cost concentrates, and the differences are structural rather than commercial.
Why does the model choice matter more than the negotiation?
Because the levers available inside a model, FUE sizing, escalation caps, bundle scope, all operate within a frame the model already set. Across the engagements advised, the deployment choice moved more money than any lever inside it.
Is RISE always the right answer?
It is always the opening answer. SAP account teams opened with RISE in nearly every cycle observed regardless of the customer infrastructure strategy. That is the account team doing its job; it becomes expensive when the opening position is treated as the shortlist.
What is the perpetual versus subscription asymmetry?
Subscription usage rights end at term end. Perpetual rights survive a dispute. A buyer holding perpetual licences who disagrees with SAP has a running system and an argument. A buyer on subscription has a deadline.
How negotiable is the conversion credit?
Highly, and it is the largest single number in most conversions. It varied by double digit percentages between comparable deals, and the variance tracked how live the buyer alternative was rather than anything about the estate.
How often is the exit actually modelled?
In fewer than half of first proposals we reviewed. That matters because on a subscription the cost of disagreement includes migrating an ERP estate under time pressure, which is the number that decides what year four feels like.
How should FUE be sized?
From measured authorisation usage rather than contracted named users. The contracted figure records what was bought and the measured one records what is used, and the gap between them is real money in the envelope.
When should indirect access be resolved?
Before it prices into a RISE envelope. Once inside the bundle it is being valued at SAP number rather than argued on its merits, which removes the ability to contest the exposure separately.
Does keeping an alternative alive actually help?
Yes, measurably. SAP negotiates hardest against the already decided, and conversion credit variance across comparable deals tracked how credible the buyer alternative remained at the point of signature.
Can the infrastructure contract be kept separate?
Under hyperscaler hosted perpetual, yes, which keeps that negotiation yours. Under RISE the infrastructure is bundled and billed at SAP rates rather than the rates you could negotiate directly, which is a cost and a control question at once.
