The S/4HANA deployment decision is a customization decision wearing a cost label
SAP offers five paths to S/4HANA, on premises, Private Cloud Edition, Public Cloud Edition, RISE, and GROW, sharing the same core product but differing on who operates it and how much you can change. Across the deployment decisions we advised, the model choice shaped cost more than the discount did, and the trade is consistent: the more SAP manages, the less you configure, and the less you negotiate later.
Prepared by Redress Compliance · August 8, 2026 · SAP advisory. Based on 40 to 50 S/4HANA deployment decisions advised 2024 to 2025.
Executive summary
Count the custom objects before the run rate chooses the model.
The common advice makes Public Cloud the default destination because it is the most standardized and cheapest to run, and it failed in our files in a specific way: complex estates steered to Public Cloud faced 20 to 40 percent rework refitting custom logic into side by side BTP extensions.
And the cheaper run rate was erased by the migration cost.
If you genuinely run standard processes, Public Cloud wins; if you do not, Private Cloud Edition protects both your code and your leverage, because PCE keeps the SAP namespace open while Public Cloud locks it.
RISE is a wrapper, and the wrapper carries a premium.
RISE is not a sixth product: it bundles PCE or Public Cloud with the HANA database, hyperscaler capacity, managed basis, and BTP credits into one subscription line, and RISE bundles priced 10 to 20 percent above the sum of the same parts bought separately on a hyperscaler.
The bundle also trades leverage, converting separate negotiations, the license, the infrastructure, the operations, into one line renegotiated as a whole, so the convenience prices twice, once in the premium and once in the lost negotiating surface.
On premises is the control path with an underestimated bill.
The perpetual license in your own data center or tenancy keeps full code freedom, your own upgrade timing, and the most negotiating leverage.
And it also keeps the HANA database as your cost to license and operate: on premises buyers underestimated the maintenance and HANA cost by 15 to 25 percent over five years.
GROW, the fifth path, packages Public Cloud for net new mid market customers and is not a route for a complex existing estate, whatever the campaign suggests.
Leverage and lock in run on the same axis as control.
The models where you hold the hyperscaler contract or the license keep the most leverage, on premises and standalone PCE with your own cloud deal; RISE bundles the leverage into one line.
And multi tenant Public Cloud holds the least, with the most standardized terms and extension only development creating the deepest lock in. Brownfield conversion preserving history and custom logic suits PCE; greenfield rebuild suits Public Cloud and GROW.
And the model you pick shapes the next decade, so the exit is weighed before the commitment, not after.
The five paths, compared
| Model | Operated by | Code freedom | Best fit |
|---|---|---|---|
| On premises | You | Full, and the HANA cost is yours | Heavy customization, maximum control |
| Private Cloud Edition | SAP, single tenant | SAP namespace open | Complex estates moving to cloud |
| Public Cloud Edition | SAP, multi tenant | Extension only, through BTP | Genuinely standard processes |
| RISE | The SAP bundle | Depends on the wrapped edition | Wrapped transformation, priced consciously |
| GROW | SAP, multi tenant | Extension only | Net new mid market, not complex estates |
The trade is consistent across the whole table. As SAP takes on more operations you gain speed and lose configuration freedom, and the negotiation surface shrinks with the code freedom: every layer SAP absorbs is a layer you no longer price separately, compare competitively, or exit independently.
Map the customization need before choosing, because the deployment decision is really a customization decision, and the estates that let the run rate decide paid for it in rework.
The three questions that decide most cases
- How much custom code do you really run? Inventory the custom objects and the processes they support: high custom load points to PCE or on premises, a clean standard footprint to Public Cloud or GROW.
- Which model keeps your negotiation leverage? Holding the hyperscaler contract or the license keeps it; the RISE bundle trades it for convenience, a trade worth making only at a price that reflects it.
- What does the exit look like? Public Cloud creates the most lock in through standardization and extension only development, on premises the least, and the model shapes the next decade.
- The migration follows the answer: brownfield conversion for PCE where history and custom logic survive, greenfield rebuild for Public Cloud and GROW, hybrid selective at higher project cost.
The S/4HANA deployment models brief
The five paths, the customization inventory method, the RISE unbundling arithmetic, and the exit terms worked end to end.
Get the white paper →The RISE arithmetic, unbundled before it signs
Pricing RISE against the sum of the parts is the exercise the bundle is designed to discourage, and it is worth an afternoon: the wrapped edition license, the HANA database, the hyperscaler capacity at your negotiated rates rather than SAP's, the managed basis at market rates.
And the BTP credits you would actually consume, compared line by line against the single RISE number, which ran 10 to 20 percent above that sum in our files.
The comparison also reveals the leverage geometry: the standalone stack renegotiates piece by piece on separate calendars, while RISE renegotiates as one line on SAP's calendar, with the year four price cliff documented in the audit defense framework's engagement data waiting at the first renewal.
The full migration decision, including the ECC deadline pressure driving it, is worked in the S/4HANA migration report, and the licensing layer underneath in the SAP licensing guide.
- Percentile standing for your exact deal size and industry, from real closed transactions
- Scenario simulation before the call: test alternative terms and see the financial impact of each
- A negotiation playbook, talking points, and a two page executive brief on day one
What we saw across deployment decisions, 2024 to 2025
Across roughly 40 to 50 S/4HANA deployment decisions we advised on between 2024 and 2025, the model choice shaped cost more than the discount did:
Custom code refitted into extensions when complex estates were steered to Public Cloud.
RISE against the same parts bought separately on a hyperscaler, before the leverage cost.
The on premises underestimate rounds out the picture: buyers who kept control underestimated the maintenance and HANA cost by 15 to 25 percent over five years, which means every path has a number its advocates leave out, the rework on Public Cloud, the premium on RISE.
And the operations bill on premises.
The honest comparison prices all three, sized to your actual custom object inventory, and then fixes the exit and true down terms before any subscription signs, because the model decision is the last moment those terms are negotiable from strength.
Your first five moves
- Inventory the custom objects and the processes they support, before any model conversation, because the count decides the case.
- Price RISE against the sum of the parts on your own hyperscaler rates, where the 10 to 20 percent premium shows itself.
- Decide who holds the hyperscaler contract deliberately, since the holder keeps the leverage.
- Model brownfield against greenfield honestly, with the 20 to 40 percent rework priced into any Public Cloud case.
- Fix exit and true down terms before signing, the last moment they negotiate from strength. The SAP practice runs the decision with you.
Frequently asked questions
What are the S/4HANA deployment models?
Five paths sharing the same core product: on premises with a perpetual license you operate; Private Cloud Edition, a single tenant SAP managed subscription with the namespace open; Public Cloud Edition, multi tenant and extension only through BTP.
RISE, a commercial wrapper bundling either edition with infrastructure and operations; and GROW, Public Cloud packaged for net new mid market customers.
What is the difference between RISE and standalone PCE?
RISE is not a separate product: it wraps PCE or Public Cloud with the HANA database, hyperscaler capacity, managed basis, and BTP credits into one subscription line, and the bundles priced 10 to 20 percent above the same parts bought separately in our files.
The standalone route keeps the hyperscaler contract, and the leverage, in your hands, renegotiating piece by piece instead of as one line.
Should we choose S/4HANA Public Cloud?
Only if your processes are genuinely standard: Public Cloud locks the SAP namespace, custom logic must rebuild as side by side BTP extensions, and complex estates steered there faced 20 to 40 percent rework that erased the cheaper run rate.
Count your custom objects first; a clean footprint points to Public Cloud, a heavy one to PCE or on premises.
What does S/4HANA on premises really cost?
More than most models assume: the perpetual license keeps maximum control and leverage, but the HANA database is yours to license and operate, and on premises buyers underestimated the maintenance and HANA cost by 15 to 25 percent over five years in our decisions.
The control is real, and so is the operations bill its advocates leave out.
Which S/4HANA model preserves negotiation leverage?
The ones where you hold the contracts: on premises or standalone PCE with your own hyperscaler deal keep license, infrastructure, and operations as separate negotiations on separate calendars.
RISE bundles them into one line renegotiated as a whole on SAP's calendar, and multi tenant Public Cloud holds the least leverage with the most standardized terms.
Is GROW with SAP right for existing SAP estates?
No: GROW packages Public Cloud Edition for net new mid market customers, extension only and standardized, and it is not a route for a complex existing estate whatever the campaign suggests.
An established estate with custom code weighs PCE against on premises, with Public Cloud reserved for the genuinely standard footprint.
