An EA renewal with a Copilot expansion, an Azure commitment, and an MCA-E migration proposal on the table at once, sequenced into one deliberate position. Three knowledge checks along the way, and 1 clip from a senior cloud advisor.
The presenter in this session is an AI generated avatar. The curriculum and guidance are real, produced by Redress Compliance analysts from our consulting engagements and market network.
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. Once in the session the frame splits and a senior cloud advisor gives the view from inside real Oracle 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 to session forty of forty, the capstone. Thirty nine sessions of pieces, and today we put them on one table at once, because that is how they actually arrive. And I want to describe the situation before I describe the method, because the situation is the lesson. A modern Microsoft proposal does not contain one decision. It contains four: a renewal, a Copilot expansion, an Azure commitment, and increasingly a suggestion that the whole relationship changes vehicle. Those four have different owners inside your organisation, different evidence, and different natural timings. And they arrive as one document with one number. Which means that if you answer it as one document, three of those decisions get made by whichever evidence happens to be loudest in the room, and the other three functions never get to contribute what they know. Taking the bundle apart is the whole job.
Five takeaways. One, the separation: four decisions arrive as one document, and answering them together is how three get decided by the fourth. Two, the sequence: right size, baseline, alternative, timing, commitment, which is the session twenty eight order applied across four workstreams rather than one. Three, the evidence: every position in this capstone rests on a document produced in an earlier session, and none of them can be produced inside the negotiation window. Four, the one page: what you actually walk in with, four decisions, four recommendations, four sources, and the walk away agreed in advance. Five, what good looks like, which is not winning every point, it is landing each of the four on its own merits with the rights secured and nothing decided by a deadline you did not set.
The situation, a realistic composite and probably a familiar one. Nine thousand seats, an EA renewing in eleven months, and a proposal that arrived early. What is on the table: a renewal at roughly nine percent above current spend, a Copilot expansion from eighteen hundred seats to the full estate, a larger three year Azure commitment, and a suggestion that everything moves to the Microsoft Customer Agreement. What you know: Copilot active use sits at thirty four percent at nine months, concentrated in four personas. Azure consumption grew twelve percent last year against a forecast of forty. Headcount is flat and one division is under strategic review. What you do not have yet: an MCA quote, a completed entitlement baseline, and a decision about that division. All three are obtainable in eleven months and none of them in the last sixty days. And the proposal is a good one, from the vendor's perspective, which is worth acknowledging honestly.
Four decisions, not one, with different owners, different evidence, and different timing. The renewal basket: owned by licence management from the entitlement baseline, settled by session twenty eight, a right sized estate priced against list. The Copilot expansion: owned by the business units from adoption telemetry, settled by session thirty eight, which asks whether the remaining roles match the ones that worked. The Azure commitment: owned by infrastructure from trailing consumption, settled by session twenty one, trailing actuals plus funded growth rather than the forecast. The vehicle: owned by finance from the three year projection, settled by session thirty seven on growth direction and the reduction scenario. And the package, owned by nobody at all, which is exactly why it should never be answered as one. A bundled proposal spans four functions and therefore has no owner, so answering it whole guarantees the loudest voice decides all four.
First check. The four items are offered as a package with a combined discount. What is your first move? A, negotiate the combined discount upward. B, ask for each of the four priced separately, because they have different owners, different evidence, and different timelines, and a package can only be accepted or refused whole. C, accept the two you want and decline the rest. D, decline the package and start again. Pause it, and as you think, ask yourself what a combined discount is actually a discount on.
The answer is B. A negotiates the size of a number whose composition you have not established, which has been the recurring error since session ten and is at its most expensive right here, because this package contains an Azure commitment sized on a forecast that already overran by a factor of three and a Copilot expansion into roles your own data says will not sustain use. C is closer and it is premature, because you cannot yet say which two you want: the Copilot expansion depends on persona analysis you have not finished and the vehicle depends on a divisional decision the business has not made. D discards a genuinely useful artefact, because the proposal tells you what the vendor wants, in what order, and how early they were willing to move, and all of that is intelligence worth keeping. Ask for four prices on four sheets, and let each be answered by the function that owns its evidence.
The sequence, five steps across eleven months. Months eleven to eight, evidence: entitlement baseline, three year seat projection, Copilot persona telemetry, and trailing Azure actuals, which is four documents with four owners, all from earlier sessions. Months eight to five, options: the MCA quote for comparison, the right sized basket priced against list, the Copilot expansion modelled by persona, and the Azure commitment built from actuals. Months five to three, positions: four recommendations on one page, the walk away agreed with the executive sponsor, escalation opened, and then table the shape before discussing rate. Months three to one, negotiation: rate at full volume, tranches for Copilot, ramp and re baseline right on Azure, and the vehicle decided or deliberately deferred. Final weeks, close: term sheet at the moment of agreement, rights diff, and every closing ask traded rather than conceded. Notice the negotiation is two months of eleven.
Guest analyst I want to finish with something a Microsoft account director said to me at the end of a renewal, because I think it is the most useful thing anybody has told me about this work. We had been on opposite sides of a fairly long negotiation, it had gone well for the customer, and afterwards we ended up talking honestly the way people sometimes do once the paperwork is signed. And I asked him what actually made a difference from where he sat. He said three things, and none of them were what I expected. The first was that they knew their own numbers, and specifically that when he questioned a seat count they answered with a report rather than an opinion, which meant he had to go back and check rather than push. The second was that they would not discuss the package as a package. He said, quite cheerfully, that the bundle is how a deal gets simple, and this customer refused to make it simple, so he had to justify four things separately to four different people who each knew their bit better than he did. And the third was timing. He said, we started talking to them in September for a July renewal, and by the time we got to the quarter that mattered to me they had already decided what they wanted and I was negotiating against a position rather than into a vacuum. Then he said the line I have quoted ever since: the customers who do well are not the aggressive ones, they are the prepared ones.
Not the aggressive ones, the prepared ones. Know your numbers, refuse the bundle, start early. Second check.
Check two. The Azure commitment in the proposal assumes forty percent growth. Your trailing twelve months grew twelve percent. What do you commit to? A, forty percent, since the transformation programme is approved. B, trailing actuals plus funded growth with owners and dates, sized to the floor, with a ramp and a mid term re baseline right requested. C, twenty six percent, splitting the difference. D, forty percent, in exchange for a deeper discount across the package. Pause it, and ask who pays for the gap between a forecast and an outcome.
The answer is B, and the answer to who pays is you, at term end, for capacity you never used. Forward models overran trailing consumption by twenty two to thirty eight percent, and more than half of initial commitments were oversized by fifteen to thirty percent. A treats an approved programme as delivered consumption, and a programme approved is not a workload migrated, which is a distinction the last two years of your own data has already demonstrated. C splits the difference between a measurement and a projection, producing a number with no basis in either. And D is the most dangerous answer on this slide, because it uses the package to trade an oversized commitment for a discount somewhere else, which is precisely what a bundled proposal is structured to enable and precisely why we separated the four decisions on the previous slide. Size to the floor, ask for the ramp, ask for the re baseline right, and let the discount land wherever it lands.
The position on one page, three properties of the document you walk in with, and there is nothing on it that came from the vendor. Every line has a source: the basket cites the entitlement baseline, the Copilot count cites adoption telemetry by persona, the Azure number cites trailing actuals, and the vehicle cites the three year projection, so nothing on the page rests on a preference or an instinct. Each decision has a recommendation rather than a range, because a page of options invites the other side to choose for you, and a recommendation with one sentence of reasoning is harder to move than a menu. And the walk away is on it, agreed with the executive sponsor before any meeting per session twenty nine, because if it is not written down before the negotiation then it will not exist during it. This page is the whole course compressed, and it works in the room precisely because it was not produced in the room.
What each decision lands on, a defensible outcome decision by decision. The renewal basket: right sized and priced per line, because quoted counts run twelve to twenty two percent above what is defensible. Copilot: expand to the four matching personas only, because thirty four percent active use is concentrated in those roles, per session thirty eight. Azure: sized to trailing actuals, ramped, with a re baseline right, because the forecast already overran by a wide margin. The vehicle: stay, and model properly next cycle, because the divisional decision is not made, which is the session twenty seven rule. And the rights: reduction, step down, price protection, and rate caps, because they outlast the rate. Notice that this outcome gives the vendor two of the things they wanted, a Copilot expansion and an Azure commitment. A good negotiation is not one where the other side gets nothing, it is one where each item is sized by evidence.
Last check of the course. Two weeks from signature the account team asks you to drop the mid term re baseline right to close. What do you do? A, drop it, the commercial terms are agreed and it is a small point. B, trade rather than concede: keep the right and give something that costs less, because a right removed at the close is the session thirty leak in its purest form. C, refuse and let the deadline pass. D, drop it in exchange for a further discount. Pause it, and ask two things: at two weeks out, whose deadline is binding, and what is that particular right worth across three years?
The answer is B. At two weeks the binding deadline is yours, which is exactly why the ask lands there rather than earlier, and the re baseline right is the single protection on an Azure commitment sized against a forecast that has already proven optimistic once in this very negotiation. A treats a right as a detail because the number is settled, which is precisely the mechanism by which well prepared deals leak value in the final fortnight. D is the version that feels like negotiating and is worse than it looks, because a discount applies to this term while the right governs what happens when consumption disappoints, and the discount will be considerably smaller than the shortfall it fails to prevent. C escalates a manageable moment into a stand off with a deadline attached. So offer something that costs them less: a reference, a case study, a longer payment term, a co terming concession. And keep the clause.
The forty session method, and what this whole course reduces to, which is three habits. Measure your own estate: the three way join, the persona mix, the adoption telemetry, the trailing actuals, because every position worth holding rests on a number you produced, and the vendor can already see most of that data, so the only real question is whether you have looked. Separate the decisions: price from vehicle, seats from consumption, rate from volume, capability from tier, because bundles are how decisions get made by the wrong evidence and taking them apart costs nothing except insistence. And start early enough to have options: two hundred and seventy days, a priced alternative, an agreed walk away, because buyers who opened early signed nine to seventeen percent below their first quote while buyers who started inside ninety days signed within three percent of it. That is the course.
Session forty, three sentences. One: four decisions arriving as one proposal is the central problem of a modern Microsoft negotiation, because a package has no owner and answering it whole means three decisions get made by the evidence belonging to the fourth. Two: the sequence is the method, nine months of evidence and options, two months of negotiation, and a close that trades rather than concedes, because what happens in the room is decided by what was produced before anybody entered it. Three: keep three habits from these forty sessions, measure your own estate, separate the decisions, and start early enough to have options, and everything else in this course is detail that follows from those three. Thank you for staying with it to session forty. Whatever your next renewal looks like, you will be the prepared one rather than the aggressive one, and that is the whole point.
The last homework, and this one is a month rather than an hour, because you are running the method on your own estate. One, find your date: your anniversary, counted back two hundred and seventy days, and if that date has already passed then start today and write down what you have given up, honestly. Two, list the four decisions genuinely on your table over the next cycle, and it is rarely fewer than three. Three, name the four evidence owners: licence management, the business units, infrastructure, and finance, with one document each and a delivery date attached. Four, write the one page: four decisions, four recommendations, four sources, and space for the walk away, drafted now and corrected as the evidence lands. Five, book the sponsor conversation about the walk away, before any vendor meeting, because everything in these forty sessions ultimately rests on that one conversation happening.
Five reads to finish on, the ones worth returning to when a real negotiation starts, all free on redress compliance dot com. First, the Microsoft EA renewal twelve month playbook, which is the timeline this capstone is built on. Second, the EA discount negotiation levers, for the seven levers and the order to pull them in. Third, the CIO level playbook on evaluating renewal proposals across EA, MCA, and CSP, which is the comparison sheet for the vehicle decision. Fourth, the Azure MACC sizing guide, for the commitment, which is the largest single number in most Microsoft deals and the least scrutinised. And fifth, the Microsoft licensing guide, which is worth re reading once a year simply to see what has moved. That is forty sessions. The knowledge checks throughout and the certification exam are there when you want to test what stuck. Thank you for your time, and good luck with the next one.