Contents
Key takeawaysWhere the savings areReclassifying named usersEngine driftDigital access documentsReviewing USMM and LAWUsing findings in a renewalWhat we have seenWhat to do nextFAQSAP license optimization comes down to seven areas, and most companies can cut 15 to 30 percent of the bill without changing a single system. The money sits in positions you prove with usage data before SAP measures you.
- Reclassify on activity. Twelve months of transaction history per user decides which Professional users can move to Limited Professional or Employee Self Service.
- Clean accounts first. Leavers, test accounts and duplicate identities are the least contested saving, because removing them needs no debate about entitlement.
- Watch the engines. Engines metered on orders, contracts, revenue, cores or documents drift past entitlement with no purchase to flag it, so check them every quarter.
- Count documents yourself. Digital access estimation tools overstate billable documents where integrations retry or duplicate, and the count sets the price.
- Review before you submit. Never send SAP a USMM or LAW result you have not checked, because a correction made afterwards becomes a concession.
- Start six to nine months out. Cleanup and reclassification take a quarter, and the result has to be finished before the renewal opens to change the price.
Where does SAP license optimization actually save money?
It saves money in seven specific places, and most companies can take 15 to 30 percent off the SAP bill with no change to their SAP systems. None of the seven needs a technical project. Each one needs usage evidence and a contract argument built on it.
I spent years on the vendor side of these conversations. The customers who paid least rarely had the best tool. They came with their own numbers, already checked, and SAP had to argue against that evidence.
| Area | What you need to prove | Where the evidence comes from |
|---|---|---|
| Named user reclassification | A user's activity fits a cheaper type than the one assigned | Twelve months of transaction history per user |
| Inactive and duplicate accounts | The account belongs to a leaver, a test, or a person counted elsewhere | Last logon dates and identity matching across systems |
| Engine measurement | You know any gap to entitlement before SAP does | Quarterly self check of each engine metric |
| Engine metric mapping | The contract wording supports a narrower count | Your order form and the definitions it references |
| Dead engines | The business stopped using a module that support still renews | Usage logs and the module owner |
| Digital access document counts | The estimate includes retries and duplicates that are not billable | Interface logs checked against the estimate |
| Measurement review | The USMM and LAW results SAP receives are correct | Your own run of both tools before submission |
Which areas should you work on first?
Start with reclassification and account cleanup. They are the least contested and need no debate about entitlement. Engines and documents come next because they need contract reading. The measurement review comes last, since that is where your findings become the number SAP sees.
Optimize the Estate First: The SAP Work That Pays for the Negotiation
How do you reclassify SAP named users to cheaper license types?
You match each user's license type to what that person actually does in the system, and you prove it with twelve months of transaction history. The activity record is the evidence, and SAP has to answer it. A job title proves nothing either way.
SAP's contractual user definitions are broad, and that breadth cuts both ways. The wording SAP uses to argue users upward in an audit is the same wording you use to argue them downward. The side holding the activity data usually wins the point.
| From | To | Typical trigger | Observed share moved |
|---|---|---|---|
| Professional | Limited Professional | Display heavy profiles, single module use | 15 to 25 percent |
| Professional | Employee Self Service | HR and time entry only activity | 5 to 10 percent |
| Any active type | Retired or test | No logon in 90 plus days | 10 to 20 percent |
| Duplicated users | Single identity | Same person across systems | 3 to 8 percent |
What evidence does a reclassification need?
You need three facts per user: which transactions they run, in which modules, and how often. Collect them across every productive system, including systems other than the one where the license sits. Most companies hold this data already and have never assembled it into a licensing view.
- ST03N. Workload statistics show the transactions each user ran. Check the retention settings first, since many systems keep less than twelve months.
- SU01 and SU10. SU01 shows each user's contractual type. SU10 changes types in bulk.
- RSUSR200. Lists users by last logon date, which is where inactive accounts appear.
- USMM and LAW. USMM measures each system, and LAW (SLAW2 on current releases) consolidates the results so one person counts once.
User type mechanics, including conversion to Full User Equivalents under RISE with SAP, are in our FUE and user counting guide. The tools are covered in our guide to USMM, LAW, SLAW and STAR.
How do you clear inactive and duplicate accounts?
Treat them as a data problem. An account with no logon in 90 days, or one person counted in two systems, needs no argument about entitlement. It only needs to be found, and no standard SAP process finds it for you.
Leavers are the largest group, because HR closes the employee record while the SAP user stays valid. Duplicates come from people holding IDs in ECC, BW and CRM under slightly different names, which LAW merges only if its matching rule catches them.
A worked example: 2,400 named users before and after
Say a company holds 2,400 named users: 1,200 Professional, 600 Limited Professional and 600 Employee Self Service. We use illustrative net prices of $3,000, $1,200 and $150 per user. Replace them with the prices on your own order forms.
| Step | Professional | Limited Professional | Self Service | Total users | Value at illustrative prices |
|---|---|---|---|---|---|
| Starting position | 1,200 | 600 | 600 | 2,400 | $4,410,000 |
| Retire 240 accounts with no logon in 90 days | 1,080 | 540 | 540 | 2,160 | $3,969,000 |
| Merge 72 duplicate identities | 1,032 | 516 | 540 | 2,088 | $3,796,200 |
| Move 186 Professional users to Limited Professional and 62 to Self Service | 784 | 702 | 602 | 2,088 | $3,284,700 |
The cleaned position is worth $1,125,300 less, about 26 percent of the named user value. Cleanup accounts for $613,800 and reclassification for $511,500, from moving 248 Professional users, or 24 percent of those left after cleanup.
RISE with SAP counts one Advanced Use user, five Core Use users or 30 Self Service users as 1 FUE. Map Professional, Limited Professional and Employee Self Service to those three in order, and the example falls from 1,340 FUE to about 944, roughly 396 FUE less in every subscription year.
Reclassifying perpetual licenses rarely produces a refund. The value appears as Professional licenses reused for new hires instead of bought, a smaller gap at the next measurement, and a lower baseline when SAP prices an S/4HANA or RISE conversion.
SAP digital access brief
What counts as a billable document, where estimates run high, and how the conversion discount works.
Get the white paper →Why do SAP engine licenses drift past entitlement?
Engines drift because they are metered on business metrics that change without any purchase. Package and engine licenses count orders, contracts, revenue, cores and documents, and each of those grows with the business. No procurement event flags the point where you pass what you bought.
Order and document metered engines drift most in growing businesses. Core metered engines drift after an infrastructure refresh, when new servers arrive with more cores than the ones they replaced.
How should you measure engines between audits?
Put engine metrics in the same quarterly self check as user counts: metric, licensed quantity, current value and its source. Our guide to SAP package and engine licenses lists where each metric is measured.
- Measure before SAP does. A gap you find yourself can be budgeted, fixed or folded into the renewal. The same gap found in an SAP audit arrives as a finding priced on SAP's terms.
- Challenge the metric mapping. How a metric is counted is often arguable, and the contract wording decides it.
- Retire dead engines. Modules the business abandoned keep renewing in support until someone raises them.
Can you stop paying support on an engine you no longer use?
Sometimes, and the route matters. SAP's current on premises support commitments let you terminate licenses and their support without repricing in named cases: products left only in customer specific maintenance, a failed implementation, insolvency, a workforce reduction of more than 10 percent over two years, and divestiture.
Outside those cases, dropping part of the installed base can trigger repricing of the licenses you keep. Give three months written notice before the end of the current term, and document why the engine qualifies. If it does not qualify, hold it as shelfware to trade in the conversion.
What does challenging a metric mapping look like?
Start with the order form and the price list version it references, since definitions changed across versions. Then ask narrow questions. Does revenue mean group revenue or that of the licensed entities? Do intercompany orders count? Each answer comes from your own paper.
How should you handle SAP digital access documents?
Count your own documents before you discuss digital access with SAP. Digital access prices indirect use per document created, so the count is the whole basis of the price. SAP counts nine document types, with financial and material documents weighted at 0.2 per line item.
The estimation tools overstate billable documents where integrations retry or duplicate. Check each of these against interface logs before any number goes into a proposal.
- Retries. An interface that resends an order after a timeout can put two documents in the estimate for one sale.
- Duplicates. Two integrations writing the same transaction produce two counts.
- Line item weighting. Financial and material documents should appear at 0.2 per line item. An estimate that counts them at full weight is wrong.
- Scope. Documents created by named users working directly in SAP are already licensed and do not belong in the count.
For order heavy integrations, digital access is often not cheaper than named users at list price, so run both options on your own volumes.
SAP has run conversion incentives such as the Digital Access Adoption Program repeatedly. Treat the discount as a recurring pattern and negotiate it inside a RISE or renewal deal, where it can be traded against other items. Our digital access guide covers the mechanics.
Should you review USMM and LAW results before sending them to SAP?
Yes, every time. Companies that reviewed their results before submission avoided compliance findings that averaged six figures in the group that did not. The measurement is the moment your position becomes visible to SAP.
A correction made before submission is simply your measurement. Made afterwards, it is a concession, and much harder to argue once the number sits in a system of record on SAP's side.
Which measurement mistakes cost the most?
- Unclassified dialog users. USMM sets them to the default type, which SAP counts as the most expensive user type.
- A weak LAW matching rule. If names differ across systems, one person is counted two or three times.
- Technical accounts set up as dialog users. Interface and batch accounts belong under system or communication user types.
- Classifying by job title. SAP can test any classification against activity, and a title will not survive that test.
What happens after the measurement, and how to answer findings, is in our SAP audit survival guide.
How do optimization findings change an SAP renewal or RISE deal?
A documented surplus turns the renewal from a talk about how much more you owe into one about how much less. It works only if the findings exist before the negotiation opens. Findings that arrive later get treated as your opinion, while findings on the table first become the baseline both sides price from.
- Run the full measurement privately, covering users, engines and documents, as a rehearsal for the one SAP will run.
- Reclassify and clean before any number reaches SAP.
- Quantify shelfware and map it against your roadmap and the RISE pitch, so you can trade it against something SAP wants to sell.
- Open with the measured position and trade the surplus for conversion credits, capped uplifts and term protections.
The end of ECC mainstream maintenance in 2027 pressures SAP too, because SAP needs conversions signed. Clean inflated ECC user counts before SAP builds a RISE proposal from them, and ask for every bundle to be priced line by line. Our 2027 strategy guide covers the timing, including extended maintenance to 2030.
What will the SAP account team say, and what should you answer?
- "RISE pricing is based on your current footprint." Agree, and hand over the cleaned measurement as that footprint.
- "Licenses you already own cannot be refunded." Accept it, and ask for the surplus to be credited against the conversion.
- "The estimation tool shows your digital access volume." Ask which interfaces it includes and whether retries were removed. Discuss price once both sides accept the count.
- "This conversion credit is only available until quarter end." SAP has repeated these incentives. Ask for the credit in the order form with a validity date after your own decision date.
Which contract terms should you ask for?
- Fixed metric definitions. Tie each engine metric to the price list version in your order form.
- A written shelfware credit. A stated value for surplus licenses against the S/4HANA or RISE subscription.
- An uplift cap. A limit on increases for the term and the first renewal. Our uplift cap benchmark shows what buyers obtain.
- A measurement review period. Time to correct results before SAP treats them as findings.
When should each step happen?
Run the measurement six to nine months before the renewal. Reclassification and cleanup take a quarter, and a finished result carries weight that a promised one does not.
| Months before renewal | What to do |
|---|---|
| 12 | Check ST03N retention and start keeping activity data. List every engine and its metric. |
| 9 | Run USMM, LAW and the engine and document checks privately. |
| 6 | Finish reclassification and cleanup. Quantify shelfware. |
| 3 | Open the negotiation with the measured baseline, digital access included. |
| 1 | Get credits, uplift caps, metric definitions and review rights in writing. |
What have we seen in SAP optimization work in 2024 and 2025?
In our SAP optimization work in 2024 and 2025, the recoverable money was in arguments far more than in reports. Three patterns recurred.
- Reclassification. Activity analysis moved 20 to 40 percent of Professional users to cheaper types once data replaced job title assumptions.
- Cleanup. Inactive and duplicate accounts held 10 to 20 percent of named user licenses before the first cleanup, usually more than the project team had estimated.
- Measurement review. Reviewing results before submission avoided the six figure average findings seen where no one checked.
Why we do not start with a SAM tool's savings number
The usual advice is to buy a software asset management tool, run the measurement and trust the dashboard's savings figure. We disagree. In our work the tool found candidates and the contract argument produced the money, because no tool knows how your order form defines revenue.
Use the tool to collect data, then build the case with usage evidence before the negotiation opens.
Unused entitlement is worth most when you can trade it against something SAP wants to sell.
Treat SAP optimization as a negotiation discipline with a measurement attached. The wider SAP library sits in the SAP knowledge hub.
What to do next
- This month. Pull twelve months of user activity data across every productive system: which transactions, which modules, how often.
- Within a quarter. Reclassify users against their activity profiles and lock or merge inactive and duplicate accounts.
- Every quarter. Measure each engine metric against entitlement, and retire engines the business no longer uses.
- Before any digital access talk. Audit the billable document count and remove retries and duplicates.
- Before every submission. Run USMM and LAW yourself and review the results before SAP sees them.
- Before the renewal opens. Land the measured position with SAP first. Our SAP license optimization service runs the measurement and the renewal with you.
Frequently asked questions
How do you optimize SAP named user licensing?
Give each user the cheapest type that covers the transactions they run, and keep the activity record as proof. Start with Professional users, where the price gap is widest, and document every downgrade so it holds if SAP tests it later.
How much can SAP user reclassification save?
In our work it moved 20 to 40 percent of Professional users to cheaper types: display heavy and single module users to Limited Professional, HR and time entry users to Employee Self Service. The value depends on your net price gap between types, so price each reclassification with your own order forms.
What makes an SAP engine license run over its entitlement?
Its metric follows business volume, so an engine bought at one level can pass entitlement years later with no order signed. Acquisitions, new sales channels and hardware refreshes are the usual causes, and each should trigger an engine check.
Is SAP digital access cheaper than named users for indirect use?
For order heavy integrations it is often not cheaper at list price. The answer depends on a scoped document count and the conversion discount you negotiate, so compare both options on your own interface volumes before you sign.
Should we submit USMM and LAW results as soon as they come out?
No. Run both tools yourself, fix classification and duplicate errors, then send the result. Keep a record of each correction and its evidence, so you can explain any difference if SAP compares your submission with an earlier run.
When should SAP license optimization start before a renewal?
Start about nine months out and finish by six. That leaves a quarter for cleanup and reclassification and time to quantify shelfware before SAP sends its proposal. Start later and SAP can treat your findings as opinions instead of the agreed baseline.
Do SAM tools deliver SAP license savings?
They find candidates and collect data. The savings come from contract positions argued with usage evidence: whether an activity profile justifies Professional, how an engine metric is counted under your wording, and what qualifies as a billable document.
Does reclassifying SAP users reduce support fees?
Not directly on a perpetual contract, because the licenses stay in your installed base and support is charged on them. The benefit comes from licenses reused for new hires, a smaller compliance gap and a lower baseline for S/4HANA or RISE pricing. A workforce reduction of more than 10 percent over two years is one case where SAP now allows license termination without repricing.