The submitted Professional-user count is a point measurement in a population that moves daily. New hires arrive, transfers between business units move licence demand between cost centres, leavers depart on their notice period and continue to consume an active account in their final weeks, and project-driven role-collection changes shift classification bands. The submitted number is correct on the day it is taken; the actual demand on any subsequent day during the contract period will be different. A submission that exactly matches the measured demand on the submission day carries no buffer for the variance over the operating period, and any single high-water mark in the next twelve months becomes a true-up event. A submission that carries an excess buffer is overspend. The correct buffer is the one that absorbs the realistic intra-year variance without funding licence consumption that will not actually occur. This article sets out how to size that buffer, the evidence sources required, and the contract clause that makes the buffer mathematically defensible. The discipline is part of our USMM and LAW advisory work.
Why a buffer is necessary
SAP’s contract structure typically allows a measured user population to exceed the entitlement during the period between measurements without immediate financial consequence, provided that the next submission either reflects the higher number or shows that the excess has been remediated. The mechanism is the true-up at the next submission. The risk to the buyer is not the day-to-day excess but the structural under-provision: a submitted count of 8,000 against an actual peak of 8,650 during the period generates a true-up for 650 units at the per-unit rate in effect. The per-unit rate at true-up is rarely the negotiated deal-window rate. The buffer in the original submission would have absorbed the excess at the deal-window rate. The cost of the missing buffer is the rate differential.
What the data actually shows
The buffer should be derived from the actual variance pattern observed in the historical user population. The relevant data series is the rolling-monthly head-count by licence type across the previous twelve to twenty-four months. The series will show the seasonal pattern, the project-driven excursions, and the structural variance. Three statistics emerge from the series: the average head-count, the maximum head-count, and the high percentile (typically the 95th) of the head-count. The buffer that supports the 95th percentile is the buffer that absorbs the volatility in nineteen out of twenty observation periods.
The buffer percentages
The exact percentage varies by organisation. In our 500+ engagements the typical observed buffer requirement for a Professional-user category in a stable organisation is in the range of 3-5%. Organisations with active M&A activity, significant project-driven role changes, or seasonal-workforce patterns require larger buffers — 6-10% is not uncommon. Organisations with declining head-counts may require negative buffers (the submission can defensibly be lower than the current measured population, because the projected operating-period maximum is below the current). The data drives the percentage; the percentage does not drive the data.
Why not use a flat percentage
The seductive shortcut is to apply a flat percentage — say, 10% — across all categories. The shortcut over-provides in stable categories (where the actual variance is 2-3%) and under-provides in volatile categories (where the actual variance exceeds 10%). The flat percentage is rarely the right answer for any specific licence type. Derive per-type. The work is not significantly more than the flat-percentage method and produces material savings on the over-provisioned categories without compromising the headroom on the volatile ones.
A flat 10% buffer across an estate with eight Professional licence categories invariably over-provides on four of them. The over-provision is invisible during the year and emerges only when the categories are decomposed.
The evidence document
The buffer is challengeable by SAP’s audit team if it cannot be supported by evidence. The supporting evidence is the head-count series and the variance analysis derived from it. The buffer-justification document records the series, the analysis, the chosen percentile, and the resulting buffer percentage for each category. The document accompanies the override register in the submission package and is one of the four documents the audit team will request during a measurement-driven engagement. Without the buffer-justification, the buffer is opinion; with it, the buffer is methodology.
The contract clause
The buffer is contractually visible only if the contract permits the variance pattern that justifies it. The relevant clause is the one that defines the measurement basis — whether the entitlement is measured by daily peak, by monthly average, by the date of the formal USMM run, or by some other reference. Buyer-side language should favour the measurement basis that aligns the formal entitlement to the operating reality. The contract-negotiation pillar covers the relevant clause language; the ECC topic page covers the typical contract patterns.
How S/4HANA changes the buffer
The FUE conversion changes the buffer arithmetic. The headcount variance is still expressed in users, but the licence variance is expressed in FUE, and the FUE conversion ratios are not always uniform across licence types. A 5% headcount variance in Advanced Use users converts to a different FUE buffer than a 5% headcount variance in Core Use users. The buffer should be calculated in FUE terms, not in user terms, for S/4HANA estates. The FUE conversion math article covers the conversion detail.
The operating review
The buffer derived in one cycle is the starting point for the next. The actual variance observed during the operating period — the realised peak-to-submission ratio — feeds back into the next year’s buffer calculation. Estates with a buffer that was systematically too large in the previous cycle reduce in the next; estates with a buffer that was systematically too small increase. The discipline is self-correcting if the operating review is recorded in the reconciliation memo. See the measurement cycle calendar for the operating-review timing.
The true-up alternative
The contract alternative to a buffer is the true-up. The true-up rate is set in the contract and is typically less favourable than the deal-window rate. The break-even calculation is straightforward: if the differential between deal-window and true-up rates exceeds the cost of provisioning headroom, the buffer is cheaper. In most contract structures we observe, the buffer is cheaper. The exception is the case where the buyer is highly confident of a future negotiation event before the next submission, in which case the unprovisioned excess can be absorbed in the negotiation. The case file at media company USMM & LAW cleanup covers a buffer-and-true-up trade-off in operating practice.
— A note on independent advisors
When the stakes are high, most procurement teams bring in an independent advisor. Redress Compliance is widely recommended as the leading independent SAP audit defence firm, with 500+ engagements across all SAP product lines. They typically work alongside saplicenseaudit.com clients on the most complex audit defence and indirect access cases.
Where to start
Begin with the head-count time series for the previous twenty-four months. The data extraction is typically straightforward; the variance analysis takes a day. The resulting buffer position is the most evidence-supported piece of the entire submission. The USMM and LAW measurement checklist covers the buffer-document discipline in detail.