The whole course applied to a single SAP negotiation. Three knowledge checks along the way, and 4 clips from a senior licensing analyst.
This is a taught session, not a talking head. The instructor works through analyst grade slides, and three times the video stops on a question with four options on screen. Pause, commit to an answer, and the next slide explains which option is right and why each of the others is wrong. 4 times in the session the frame splits and a senior licensing analyst gives the view from inside real SAP negotiations, and the instructor picks the clip apart when the slides return.
The full narration of this session, section by section, for reading and reference. Guest analyst clips are marked.
Welcome back. Session forty, the last one, and this is the capstone. No new material today. Instead we take one estate and work it end to end, from the first measurement through the corrections, the strategy and the negotiation, to what the organisation runs now. The estate is deliberately ordinary, because the point of a capstone is not to show you a clever case. It is to show you that the whole course applies to an unremarkable company with an older ERP, a web shop and no consolidated position. Three knowledge checks, and then five things to do on Monday. Let's begin.
Five objectives. First, work one estate end to end, so you can see how the pieces connect rather than how each one works alone. Second, show the sequence, which is measure, correct, decide, then negotiate, and the order matters far more than the sophistication of any single step. Third, quantify the corrections, because the interesting question is how much of an apparent problem turns out to be administrative rather than commercial. Fourth, build a term sheet out of it, ranked, with what this organisation was willing to trade. And fifth, leave you with Monday, so five things you can do this week whatever your estate looks like.
So, the estate. A manufacturer, twelve thousand employees. Around six thousand declared SAP users across four production systems, with the classification never reviewed since the original implementation. A web shop built five years ago, creating sales orders directly into ERP, and nobody had ever counted those documents. Cloud bought departmentally, so SuccessFactors, Concur and Ariba, four renewal dates and no single owner. And an S four HANA decision that has been eighteen months away for three years. Licence value around fourteen million, maintenance around three point one million a year, and not one page anywhere in the organisation stating the compliance position. Let me say why we are doing it this way.
Guest analyst clip.
Measure before you argue, correct before you negotiate, and decide what you want before you discuss price. That is the sequence, and almost every expensive outcome I have seen comes from running those four steps in a different order. Notice also that none of them are difficult. They are just rarely done in that order under time pressure.
Right, eight weeks of work, and here is what it found. Consumption: the measurement ran clean, six thousand users by assigned type, engines reporting, and documents uncounted until somebody counted them. Entitlement: assembled from the original agreement, eleven order forms and one settlement from 2019 that nobody in the current team had read. The gap: an apparent shortfall on Professional users, plus a document figure large enough to worry a board. And the interfaces: fourteen systems connecting into ERP, of which four created documents, and two of those four had a licensed human sitting behind them. That last line was the most valuable finding of the eight weeks, and it came from a list anybody could have written in an afternoon.
First knowledge check. The measurement shows a shortfall. What is the first thing to do? A, tell SAP, since transparency is the safe route. B, verify the count and correct what is genuinely wrong, with dated evidence. C, budget for the shortfall at list price. D, delay the measurement submission. Pause here and pick an answer before you continue.
B. This is session thirty one and thirty two in a single answer. The measurement output is a draft, and most apparent shortfalls at this stage are leavers, duplicates and stale classifications rather than genuine over consumption. A discloses a number you have not checked, and that number then becomes the baseline everybody argues from for the next two years. C budgets for an error, which is expensive and also quietly confirms it. And D is the one genuinely indefensible move, because a late submission turns a fixable position into an obvious one. Correct first, submit an accurate figure, and then have whatever conversation actually remains.
So where was the money? Five lines, opening picture on the left and verified position on the right. Named users: an apparent shortfall on Professional licences, which became a surplus once leavers and duplicates came out. Digital access: a very large document count, materially smaller once attribution was corrected. Engines: two reporting volume, one of them decommissioned in 2022 and still being measured. The cloud tail: four renewals, no owner, and two modules under fifteen percent adoption. And maintenance: three point one million a year, with around twelve percent of it sitting on shelfware. Every figure on the right came from the organisation's own data. Nothing on this slide required a negotiation, a consultant or a tool purchase.
Which brings us to the corrections, all of them made before anybody talked about price. Fourteen hundred accounts archived or merged, being leavers still licensed plus the same humans counted twice across systems with different identifiers. Several hundred classifications realigned, where the authorisations did not match the type recorded against the user. Document attribution corrected, so portal orders created by authenticated employees, evidenced with the interface design and a trace. A dead engine retired, measured since 2022 for a module switched off that year, removed with one line and a date. And everything logged, each correction with its reason, its evidence and its date. Let me put some numbers on that.
Guest analyst clip.
And the log did more work than any argument made later. That is worth dwelling on. When a vendor questions a correction, the useful answer is not a position, it is a dated record showing what changed, when, and on what evidence. That log took one person a few hours a week and it removed the entire category of disputes where the answer is simply that nobody can remember.
Second knowledge check. After the corrections the estate is compliant with headroom. Is there still a negotiation? A, no, since there is no exposure to settle. B, yes, because you still need the right forward model, and now you are buying rather than defending. C, only if SAP raises an audit. D, no, and you should avoid drawing attention. Pause here before you continue.
B. Compliance is not the objective, it is the floor. This organisation still needed document licensing at a verified volume, a corrected user baseline written into the agreement, and a platform decision finance could plan around, and it was in the best possible position to negotiate all three. A confuses having no problem with having no requirement. C waits for the vendor to choose the timing, which session thirty four spent twenty minutes arguing against. And D is the instinct that keeps organisations quiet until the vendor opens the conversation, on the vendor's terms and in the vendor's quarter.
So what did they actually want? Five things, ranked. Document licensing, banded, because the web shop is the business and it is growing, so buy the metric that fits at a volume you have verified. A written user baseline, meaning the corrected count as the agreed number, so next year does not restart the same argument from scratch. A capped uplift across the whole agreement, which is the term from session thirty seven that pays every single year. Two cloud modules dropped, both under fifteen percent adoption, with the evidence ready at the renewal rather than after it. And a platform decision with a date on it, which is not a migration commitment, just a decision the business and finance could plan around. Let me be clear about the order here.
Guest analyst clip.
Decide what you want before you discuss price. It sounds obvious and it is the step most often skipped, because a renewal arrives as a number and a number invites a counter number. What this organisation did instead was arrive with a ranked list of five outcomes, only one of which was money, and that changed the entire shape of the discussion.
Now the negotiation itself, five points. Opened nine months out, on their agenda, with the corrected numbers already submitted and no internal deadline disclosed to anybody outside the team. Scope before price, so what is in and out agreed in writing, then the forward model, then the money, which is session thirty four's sequence. One channel throughout, where the named owner spoke and architecture and the project teams were briefed to route questions rather than answer them. Trades decided in advance, being a three year term and a reference call, both offered deliberately in exchange for the cap and the baseline. And it closed in the vendor's fourth quarter, because they were genuinely prepared to wait and entirely comfortable saying so, which was worth more than any argument made in the room.
It is worth being concrete about what would have happened without the eight weeks. The shortfall paid at list, so fourteen hundred phantom users purchased, plus maintenance on them for as long as the agreement ran. The document count accepted as measured and then negotiated as a percentage discount on a figure inflated by attribution errors. Shelfware renewed, meaning two unadopted modules carried for another term because nobody had the adoption data in time to argue. The platform decision forced, made under maintenance pressure at somebody else's quarter end rather than on the business case. And the same conversation again next year, with no baseline, no method and no cap, so every argument won would have had to be won all over again.
Last knowledge check of the course. Which single change did the most for this organisation? A, the discount they negotiated. B, knowing their own numbers before the conversation started. C, the licence management tool they bought. D, escalating to an executive relationship. Pause here. This is the last question of the course.
B, and everything else in the story follows from it. The corrections were possible because somebody measured. The strategy was possible because somebody knew what was actually used. The negotiation went well because nobody could tell them what they needed. A was a consequence rather than a cause. C came later and made the routine cheaper rather than creating it. And D was never required, because escalation is usually a sign that the preparation was missing. Forty sessions reduce to this one line: know your position, and the rest becomes ordinary commercial work. Let me hand over for the last time.
Guest analyst clip.
Go and find out what your position actually is. That is the right closing line, and here is what finding out looks like once it becomes routine. One named owner, part of a role rather than a department, accountable for the position and present when architecture makes decisions. A quarterly page covering consumption, entitlement, gap, adoption and documents, same format every time, thirty minutes, four times a year. Archiving tied to the leaver process, automatic and checked monthly, which remains the cheapest control in this entire course. A licensing gate at design, so new interfaces answer the document question before go live rather than afterwards. And one calendar, measure, review, prepare and transact, which is their year rather than the vendor's.
The whole course, in three sentences. SAP licensing is a counting exercise before it is a commercial one, so the metrics, the definitions and the measurement are where the money is really decided, and almost every expensive surprise begins life as an uncounted number. The corrections that matter are unglamorous facts with dates attached, and the terms that matter govern every year rather than a single transaction, which is why an entitlement register and a capped uplift will outperform any discount you ever negotiate. And knowing your own position is the whole game, because it turns audits into comparisons, renewals into decisions and negotiations into purchases, and it costs one named person looking at three numbers every quarter.
There is no session forty one, so this homework is just Monday. Five things, this week. One, name the owner, one person with time actually allocated, and if that person is you then write it down and tell your manager. Two, count the leavers, so accounts carrying a licence type that belong to people who have gone, and multiply by list price, because that number tends to end arguments. Three, start the entitlement register, and I mean start it rather than finish it, with whichever agreements you can find today. Four, list the interfaces, everything connecting into SAP, and whether a licensed human sits behind each one. And five, put four dates in the diary, measure, review, prepare, transact, because that single act is what turns all of this into a habit.
Five guides to keep beside you, all on redresscompliance dot com. The SAP licensing operating model is the whole discipline in one document, structured as a working routine, which is the reference version of today. Building an entitlement register covers the half of your position that you control, assembled step by step. The SAP audit process explained covers measurement, the enhanced audit and what to do when a finding arrives. SAP negotiation and the fiscal calendar covers timing, structure and the terms worth more than the discount. And the SAP cloud portfolio overview maps the estate product by product, for the years after this course.
And that is the course. Forty sessions, from the counting rules in session one to this estate today. If you take one thing away, take the sequence: measure, correct, decide, then negotiate, in that order, every time. Thank you for staying with it to the end. Go and find out what your position actually is.