Third-party SAP benchmarks span 140 to 360 euros per FUE per month, so a raw number is worthless at the deal desk until you normalise it to net ACV per FUE at your volume band and term

SAP publishes no list price for Cloud ERP Private, which means every per-FUE figure in circulation is a third-party estimate rather than a rate card. That gap is exactly what the deal desk exploits when it asks where your number came from. The fix is not a better source, it is a structured, normalised, indexed statement that survives challenge without exposing a single customer contract.

Prepared by Redress Compliance · August 29, 2026 · SAP advisory. RISE, Cloud ERP Private and ECC conversion engagements, 2024 to 2026.

The 140 to 360 euro spread is not evidence of a chaotic market. It is evidence that four pricing variables were stripped out of the published figures before they reached you. Edition (Base, Premium, Premium Plus) changes the bundle before anyone talks price.

Volume band moves the unit rate across five tiers running from 60 to 550 FUE up through 25,000 plus, and SAP re-cut those tiers when RISE became Cloud ERP Private, so some bands now carry a higher list per FUE than the RISE Premium equivalent. Term (three or five years) shifts the rate again.

Bundle scope decides how many SKUs sit under the same denominator. Walk into the deal desk with "I have seen 180" and the AE has four ways to say your number describes a different deal, all of them true. The fix is not a better source.

It is a stack you run before the number leaves your building: strip excluded SKUs, normalise BTP credits out using SAP's own 1 percent of net ACV formula (floored at 10,000 euros, capped at 20,000), convert every user type to FUE at the published ratios, then express one figure only.

Net ACV per FUE per year.

That figure is the only one that survives a challenge, because every distortion has already been removed in front of the deal desk rather than behind it.

Run the bundle adjustment first. Price the incremental SKUs at their standalone equivalents, subtract that from current ACV, and only then compute the per-FUE delta.

Most buyers who do this find the real increase is smaller than the headline, which sounds like bad news until you realise it tells you exactly which SKUs to hand back for a rate concession you can defend. See our work on SAP discount bands by spend tier for where the normalised figure should land.

Once the number is clean, you choose how much of it to reveal. There are three formats and they trade credibility against exposure in different directions. Structured is a comparison matrix: your quote set against modelled cohorts by volume band and term, no customer names, no contract extracts.

It is the most persuasive on paper and the most dangerous in the room, because it invites the deal desk to attack the cohort composition rather than the price. Per-unit normalised is the single figure: net ACV per FUE per year, plus the band and term it applies to.

Its power comes from what it removes. SAP publishes no list price for Cloud ERP Private, so the discount percentage the AE keeps quoting is theatre off a number SAP controls. A per-FUE figure denies them that ground entirely.

Indexed is a statement of position, not a number: "this quote sits above the third quartile for our band and term," or "we are aware of at least four comparable deals below this rate." It discloses nothing, cannot be reverse-engineered to a source, and is the format to lead with.

The reason is tactical rather than evidential. An indexed statement forces SAP to argue that your position is wrong without being able to say what right looks like, since doing so means putting their own book on the table.

That is a conversation the deal desk will not have, which is precisely why it works. The same logic underpins proving an AWS discount benchmark without naming the source.

Sequence matters more than format quality. Lead indexed. Hold per-unit normalised for the second meeting, when the AE has escalated and needs something concrete to carry upstairs.

Release the structured matrix only if you are being told your position is fabricated, and even then release the axis labels, not the cells. Each escalation costs you exposure and buys you specificity, so spend it deliberately.

In our experience across SAP negotiations, buyers who lead with the indexed statement and never move past per-unit normalised close closer to their target than buyers who open with the full matrix. Volunteering detail signals you need the benchmark to be believed.

Withholding it signals you already know where the deal lands.

Start with what SAP has actually built. There is no published list price for Cloud ERP Private. Its own pricing and packaging material carries no rates. That is not an oversight, it is the operating model.

When no list exists in the public record, the only price signal a buyer can hold is the one SAP hands over in the quote, and the only vocabulary available is percentage off. SAP owns the numerator and the denominator.

It sets the list per FUE in your specific quote, then sets the discount off that list, then congratulates you on a number it authored end to end. Twenty five years across the table from this vendor and I have never seen a cleaner information monopoly in enterprise software.

A normalised net ACV per FUE figure breaks it in one move.

Once you divide total committed annual value by contracted FUE count, after stripping the adjacent SKUs that never belonged in the comparison (Business Network beyond basic, SuccessFactors, Concur, Ariba advanced, Commerce Cloud.

Country payroll) and after normalising the BTP credit line, which is formulaic anyway at one percent of net ACV floored at ten thousand and capped at twenty thousand euros, you are holding an absolute number.

Not a percentage. An absolute. And an absolute cannot be argued with by adjusting the list.

This is precisely why the deal desk attacks provenance and not arithmetic.

Watch the sequence in any live negotiation: the moment you say two hundred and forty euros per FUE per month at your volume band and term, the first question is never "how did you calculate that." It is "where did you get that." The distinction matters enormously.

Attacking your source costs SAP nothing. Attacking your method would require SAP to reference its own cohort data, and the instant it does that it has published, in the room, what deals shaped like yours actually sign at. No account executive has authority to do that and no deal desk wants to.

So the vendor stays on provenance forever, because provenance is the only front where it can fight without disarming itself.

The correct buyer response follows directly. Do not defend the source. Restate the method. Say: this is net committed value divided by contracted FUE, adjacent SKUs excluded, BTP normalised, at the 551 to 4,000 band on a five year term. Then invite correction on any input.

Nine times out of ten the desk will not correct a single line, because correcting a line means confirming the rest.

Our own work on why evidence beats anecdote at the table lands on the same point from a different direction: the buyer who has done the normalisation work is the only party in the room able to show their working, and showing your working is a form of leverage that no rebuttal script handles well.

There is a consequence here that catches sophisticated buyers, and it is expensive. SAP re-cut the volume tiers when it retired RISE Premium and moved to Cloud ERP Private. For some bands, list per FUE went up. DSAG's own licensing board member confirmed the FUE costs more than the old Premium unit.

Gartner's read is blunt: customers negotiating on discount now need more aggressive discounts simply to reach the same net price per FUE they held last cycle.

So the buyer who walks in proud of holding fifty five percent, the same fifty five percent they signed three years ago, may be paying materially more per unit while believing they defended the line.

Percentage discipline without unit discipline is not discipline, it is theatre performed against a moving denominator.

That is the whole asymmetry. SAP controls the list, so it controls the percentage. It does not control your net ACV per FUE, because that number is computed from your own committed spend and your own contracted units. The only party who can refute it is the one who will not open the book.

Price the position accordingly and stop trading in percentages that SAP writes for you.

Expect four moves, in roughly this order, and expect the sequencing to be deliberate rather than reactive. The opening anchor almost always arrives before you speak: the AE cites a list per FUE you cannot verify, then presents a "special" discount as evidence of partnership commitment.

Everything after that is designed to keep the conversation inside a frame SAP authored. Your job is to refuse the frame four times without raising your voice.

The mistake most buyers make with a good benchmark is spending all of it on the headline rate. The deal desk defends the per-FUE number hard because it is the metric SAP reports internally and the one that sets the renewal baseline. It defends the escalation clause far less.

A flat per-FUE price for the full term, or a 2 percent hard cap tied to CPI, is frequently grantable at account executive level without escalation, and it is worth more than 5 points of upfront discount on a five-year deal.

Run the arithmetic in front of them: standard SAP cloud paper carries a 3 to 5 percent annual uplift after year one, and at a 4 percent compound rate across five years that is roughly 22 percent on top of year-one pricing.

Anything above 3.5 percent is aggressive against a 2.5 to 3 percent market standard. Above 4 percent, SAP is pricing for a buyer who will not check. Our read on what uplift caps buyers are actually signing sets the range you should be quoting back.

Then work the seven surfaces where price rises without a renegotiation: the annual uplift, the FUE recategorisation clause, the BTP overage rate, the Digital Access true-up rate, the hyperscaler pass-through, support escalation, and renewal pricing.

The FUE ratchet is the one that quietly costs the most. Standard language lets your count go up and never down, so 50 Named Users added in year two become your permanent floor. Insist on a true-down mechanism that mirrors the true-up, at minimum an annual reset to actual consumption.

Recategorisation is the same problem in a different shape: SAP gets latitude to reclassify users against criteria the contract never defines. Fix it with documented criteria, a quarterly review cadence, and reclassification only by mutual agreement.

On Enterprise Support, 22 percent is already the high end. Competitive deals close at 18 to 20 percent. If you are being quoted 24 percent or more, you are being tested, not priced.

One more item most buyers miss: the ECC conversion credit runs three years and then stops. Year four renews at full subscription price with your negotiated discount, if any, applied to a base you never benchmarked.

Model that year-four number now, put a renewal cap in writing, and treat the credit as a timing device rather than a discount.

Where the sources agree, they agree firmly. There is no published list price for Cloud ERP Private, so every circulating per-FUE figure is a third-party estimate.

Discount bands run 30 to 65 percent off first proposal, with the mid-market cluster at 300 to 500 FUE landing 40 to 50 percent and larger enterprises above 500 FUE reaching 50 to 65 percent.

Uplift market standard sits at 2.5 to 3 percent, and anything above 4 percent is a signal about how SAP reads your negotiating capability, not a signal about cost.

Deals in the 1 to 5 million euro range reach 40 to 50 percent with basic competitive positioning and quarterly timing; 10 million euro plus deals reach 50 to 60 percent, but require C-level SAP approval, which changes your timeline more than your price.

Our spend-tier discount band work holds up against every source in this set.

The conflict is entirely in the per-FUE numbers themselves: one source cites 220 to 280 euros list, another 140 to 220 dollars, a third 190 to 360 dollars. That spread is not sloppiness, it is the finding.

DSAG has publicly confirmed the FUE metric costs more than the old RISE Premium unit, and Gartner has noted the volume tier re-cut raised list per FUE in some bands while nearly doubling the bundled SKU count.

Different editions, different bands, different bundle contents, all reported as one number. This is precisely why you build an internal baseline first, normalise to net ACV per FUE within your own volume band and term, and treat published ranges as a sanity check rather than an anchor.

Related advisory: the full topic guide.

Can I show a SAP account executive another customer's contract as proof of a benchmark?

No, and you should not want to. Disclosing a third party's contract breaches its confidentiality terms and immediately shifts the negotiation from your pricing to your conduct.

An indexed statement ("we have visibility of multiple comparable deals in the 551 to 4,000 FUE band at a materially lower net ACV per FUE") carries the same pressure with none of the exposure, and SAP cannot disprove it without disclosing its own book.

Why do published SAP per-FUE benchmarks disagree so widely?

Because they measure different things. Published figures range from roughly 140 to 360 per FUE per month, but none of them consistently state the edition (Base, Premium or Premium Plus), the FUE volume band (five tiers from 60 to 550 up to 25,000 plus), the term, or which SKUs are inside the bundle.

Once you normalise those four variables, the spread narrows sharply and the comparison becomes usable.

What is the single most defensible way to state a SAP benchmark?

Net annual contract value per FUE per year, at your volume band and term, with excluded SKUs stripped and BTP credits normalised out using the 1 percent of net ACV formula.

This removes discount percentage from the conversation entirely, which matters because SAP publishes no list price for the private edition and therefore controls the denominator in any percentage-based discussion.

How does SAP typically respond when a buyer produces benchmark evidence?

Four scripted rebuttals: attack the provenance, assert your deal is not comparable on bundle or geography, reframe to total value citing the near-doubled bundled SKU count in Cloud ERP Private, and apply quarter-end timing pressure so you cannot verify anything.

The correct response is to restate your method rather than defend your source, and to note that SAP cannot call your figure wrong without producing a published rate it does not have.

Which SKUs should I strip before comparing per-FUE pricing?

Business Network basic, SuccessFactors HCM, Concur, Ariba functionality beyond basic Business Network, Commerce Cloud, Customer Data Cloud, and country payroll modules all sit outside the core bundle.

Include any of these in the denominator and your per-FUE figure is inflated against a peer who priced only the core package. Normalise BTP credits out separately, since they are formula-driven at 1 percent of net ACV, floored at 10,000 euros and capped at 20,000.

Is a benchmark worth more spent on discount or on the uplift clause?

Usually the uplift clause. A standard 3 to 5 percent annual escalation compounds to roughly 22 percent on year-one pricing across a five-year term at 4 percent, which typically exceeds what an extra 5 points of upfront discount is worth.

A flat per-FUE price for the term or a 2 percent cap tied to CPI is often grantable at account executive level without deal-desk escalation, making it the cheaper concession to win.

Do I need an external benchmark at all, or can I build one internally?

Build the internal baseline first. Your prior ECC spend, your actual user distribution mapped against the FUE conversion ratios (1 Advanced, 5 Core, or 30 Self-Service per FUE, 2 FUEs per developer), and your BTP credit calculation are facts SAP cannot dispute.

External benchmarks are then used to test whether your baseline sits in the right band, not to originate the number.