Discounting, term alignment with the EA or MCA, ramps, pilots, and the traps in a product Microsoft is pushing hard. 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 back, session nineteen of forty. The last three sessions built the picture: what the Copilot subscription licenses, what the metered layer costs, and what an agent estate does to a bill. Today we turn all of that into a negotiation. And I want to open with the claim that organises this session, because it runs against instinct. The rate is the easy part. Almost every buyer I meet arrives at a Copilot conversation with a target discount in mind and no position at all on volume, term, or the metered rates, and that is precisely backwards. The rate is genuinely negotiable and it is the least valuable of the things on the table, because a good rate applied to too many seats for too long is still the wrong deal. What you are really negotiating is the shape of a commitment, and the shape is worth several times the percentage.
Five takeaways. One, what is negotiable: more than the rate. Volume shape, term alignment, ramp thresholds, the metered rates from session seventeen, and enablement funding are all available if you ask early enough. Two, the discount reality: median final discount landed six to nine percentage points above what the account team first flagged as achievable, across sixty to eighty EA renewals, so the first number is not the number. Three, the shape premium: phased deals that priced Copilot separately saved fifteen to thirty percent against the all in bundle, which beats any discount you will win on the rate alone. Four, the term trap: annual terms under New Commerce locked seats in twenty five to forty percent of estates, so commitment length matters as much as price. Five, the timing: Microsoft's fiscal year ends on the thirtieth of June, and sequencing against that and your own renewal is worth real money.
What you are actually negotiating, three things, and only one of them is a price. The rate: per seat against a list of around thirty dollars, genuinely negotiable, and the least valuable of the three, because a rate applied to too many seats is still too much money. Negotiate it as though for the full population, because you want the rate that volume commands. The volume shape: how many seats, when, and on what condition, and this is where the fifteen to thirty percent lives. A tranche structure with adoption thresholds is a completely different deal from the same rate applied to everybody on day one, at the same rate. And everything around it: term length and alignment, the metered rates and pool terms from session seventeen, agent governance entitlements from session eighteen, enablement funding, and the right to reduce. None of those appear on a quote unless you put them there. And the order matters: agree the shape while the rate is still open, because conceding the rate early removes your reason to keep talking.
The levers ranked by what they are worth. Volume shape, at the top: tranches released on measured adoption rather than a flat rollout, worth fifteen to thirty percent against the all in bundle. Rate: a per seat discount against list, meaningful, and final deals landed six to nine points above the first indication. Metered rate caps: message pool unit price and per agent rate fixed at signature, and the value there is not a percentage saved, it is an unbounded downside avoided, which is a different kind of win and easy to undervalue in a spreadsheet. Reduction rights: the ability to hand seats back at anniversary, which is the difference between a mistake and a permanent mistake. And enablement funding: Microsoft funded deployment and adoption support, which is real and routinely available on a product being pushed this hard. Notice that the top lever is not a price at all. The largest number available to you is a decision about how many people get a licence and when.
First check. The account team says the discount you are asking for is not achievable. How much weight does that carry? A, a lot, they know their own approval limits. B, some, but final deals landed a median six to nine percentage points above what the account team first flagged as achievable, so the first indication is a position rather than a ceiling. C, none, everything is always available. D, it depends entirely on your spend level. Pause it, and as you think, ask yourself who the person telling you this has to ask, and when their year ends.
The answer is B. Across roughly sixty to eighty Microsoft EA renewals, median final discount sat six to nine percentage points above what the account team had initially flagged as achievable. And I want to be fair about what that gap is, because it is not evidence of bad faith. The person in front of you has an approval chain behind them, and their opening position reflects what they can sign on their own authority rather than what the organisation will eventually do. A takes the first answer as the last one, which is exactly how a six to nine point gap survives untested. C is the opposite error and it damages you differently, because a buyer who believes everything is available negotiates without priorities and trades badly, usually giving away term length for rate. D matters at the margin, because spend level does affect the range, and it does not change the structural fact that first indications are conservative. Keep asking, in writing, with your alternative priced, and be patient enough to let it travel up the chain.
The traps in a product being pushed, five patterns to recognise before you meet them. The pilot that is a rollout: a pilot with a three year commitment attached is not a pilot, and the test is simple, which is what happens if it goes badly. The bundled uplift: Copilot folded into an E5 step up as a single number, which you met in session thirteen, and the answer is always to ask for them separately. The adoption promise: enablement support offered as a reason to commit wider, and my advice there is take the enablement because it is genuinely valuable, and do not let it move the seat count. The agentic demo: cowork and agents shown before base adoption is proven, from session eighteen, and a demo is evidence about the product rather than about your estate. And the expiring incentive: an offer lapsing at quarter or fiscal year end, sometimes entirely real, never a reason to sign something you would not otherwise sign. The common thread is that each one converts enthusiasm into commitment length.
Guest analyst There is a phrase I have learned to slow down on, which is, we are just doing a pilot. A financial services client used exactly that phrase with me and I asked what I always ask now, which is, what happens to the licences if the pilot fails. There was a pause on the call, and then somebody said they would check. What came back was that the pilot was fifteen hundred seats on a twelve month term with no reduction right, signed three weeks earlier at a good discount. So the honest description was not a pilot. It was a fifteen hundred seat annual purchase with an evaluation attached to it, and the evaluation had no consequences, because whatever it concluded they were paying for twelve months either way. Nobody had misrepresented anything. The word pilot had been used in a meeting, everyone had heard it as low risk, and the commercial terms had been agreed by somebody who was not in that meeting. Here is what I found most useful about it though. When we went back to the account team and explained the mismatch, they were perfectly reasonable about it. They restructured to three hundred seats with a genuine exit and the same rate held for expansion, because a real pilot that expands is worth more to them than a fake one that gets resented. The lesson I took was that the fix was available for the asking, and nobody had asked, because everybody in the room believed the word pilot was doing work that the contract was not doing.
Fifteen hundred seats and no exit, called a pilot by everybody involved. Ask what happens if it fails. Second check.
Check two. You are offered a discounted twelve month Copilot term for a two thousand seat pilot under New Commerce. What do you check first? A, the discount level against benchmarks. B, whether you can reduce seats before the twelve months are up, because annual terms locked seats in twenty five to forty percent of estates and a pilot you cannot exit is a rollout. C, whether the pilot includes training. D, whether the term aligns with the EA anniversary. Pause it. All four are worth checking, so the question is which one decides whether this is a pilot at all.
The answer is B. Annual terms under New Commerce locked seats for the term in twenty five to forty percent of the estates we reviewed, so the reduction question is not a detail, it is the question that decides whether the word pilot is accurate. If those two thousand seats cannot come down for twelve months, you have bought two thousand seats for a year and agreed to call it an experiment, which is the story you just heard. D is the strongest of the other three and genuinely important, because a term ending off cycle creates the stub period problem from session ten and hands the vendor a second negotiation you never planned for. C is worth having and changes nothing about the commitment. And A is the question everybody asks first, which is the point: a good discount on a pilot you cannot stop is a worse outcome than a modest discount on one you can. Ask about the exit before you ask about the price.
Term alignment, three ways the Copilot term meets the rest of your estate, and alignment here is not about tidiness, it decides how many negotiations you have and when. Co terminate with the agreement: Copilot ending when the EA or MCA ends means one negotiation covering the whole relationship with everything on the table at once, and that is usually the strongest position, because Copilot is the line the vendor most wants and you want it sitting beside the lines you most want. Deliberately shorter: a shorter Copilot term buys the right to re decide before the main renewal, it costs rate, and it is defensible when adoption is genuinely unproven, provided you hold a reduction right and you know the stub period cost. And accidentally misaligned, which is the common outcome: a mid year purchase creating its own anniversary, its own true up rhythm, and a second annual conversation. Nobody chooses that and a great many estates have it.
The calendar, whose year ends when. The thirtieth of June is Microsoft's fiscal year end and it is the single strongest window, and the way to use it is to be ready before it rather than during it. Quarter ends are internal targets inside the year, useful, weaker, and available four times as often. Your EA anniversary is your true up and order date, which is your own leverage point, so bring Copilot to it deliberately rather than letting it land in between. Your renewal window, from session twenty six, is where everything can be sequenced together. And the demo date, which is when the agentic demo lands well internally, and that is the moment to slow down rather than to commit. The principle underneath the whole slide: being ready before the window matters more than the window itself. A buyer with a priced alternative and an internal approval already in hand can move inside a vendor's deadline. A buyer who starts preparing when the deadline is announced is being timed by somebody else.
Last check, and this is a real situation rather than a constructed one. Your best Copilot offer expires on the thirtieth of June. Your evidence on adoption will not exist until September. What do you do? A, sign in June, the incentive is worth more than the evidence. B, take the rate in June on a small committed tranche, with the remaining volume priced at the same rate and released after the September evidence. C, wait until September and accept a worse rate. D, ask for the offer to be extended to September. Pause it, and as you think, notice that you can have both the rate and the evidence, so work out what has to be split to get them.
The answer is B, and the apparent conflict only exists if rate and volume are treated as a single decision. Split them and both are available: commit a tranche now at the fiscal year end rate, hold that rate for the remaining volume in writing, and release it when the adoption evidence lands in September. That is the session sixteen structure applied to a calendar problem, and account teams accept it more readily in June than at any other point in the year, because it books something now, which is what June is about for them. A pays for the rate with a commitment you cannot evaluate. C treats the vendor's deadline as though it binds you, which it does not, and it also gives up something real for no reason. D is worth asking for and it is a weaker version of B, because an extended deadline still leaves you deciding the entire volume in one go. Ask for the split first, and keep the extension as a fallback.
The Copilot negotiation method, three moves, and the order is the method. One, table the shape first: bring your tranche plan and adoption thresholds to the first meeting, before any discussion of rate. That reframes the negotiation from how much per seat into how many seats and when, which is the conversation where your numbers are strongest and theirs are weakest. Two, negotiate the rate at full volume: price as though for the whole population so the rate reflects the volume the vendor is hoping for, then commit only the first tranche, and hold the rate for the term in writing, because a rate that lapses with the tranche is not a rate you won. Three, close the surrounding terms: reduction rights, metered rate caps, agent governance entitlements, enablement funding, all of which are cheap to concede once the headline is agreed and impossible to add afterwards. And ask for the reduction right explicitly whatever the term length.
Session nineteen, three sentences. One: you are negotiating three things and only one of them is a price, so table the volume shape first, because a tranche structure is worth fifteen to thirty percent against the all in bundle and no rate concession competes with that. Two: first indications are positions rather than ceilings, with final deals landing a median six to nine points above what was originally flagged as achievable, so keep asking, in writing, with your alternative priced. Three: check the exit before the price, align the term deliberately, and use the thirtieth of June by being ready before it rather than being timed by it. Next session closes module four with the thing that makes all of this repeatable, which is the evidence.
Homework, about an hour, and this week you write the position. One, write the tranche plan from session sixteen: first tranche named, thresholds for the rest, one page, and this is what you table first in any meeting. Two, find your dates: your EA anniversary, your renewal window, and the thirtieth of June, all on one calendar so you can see what collides. Three, check your current exit: if you already hold Copilot seats, can you reduce them at anniversary, and find the clause rather than the recollection, because those two differ more often than people expect. Four, list the surrounding terms: reduction rights, metered rate caps, agent governance, enablement funding, four lines ready to raise when the headline is close. Five, price your alternative: what you do if you buy nothing this cycle, because a buyer who has priced that is a different negotiator from one who has not.
Five reads before next session, all free on redress compliance dot com. First, negotiating Microsoft generative AI contracts, which is the full buyer side playbook for the AI lines. Second, benchmarking Microsoft EA discounts, where the six to nine point gap between first indication and final deal comes from. Third, key leverage points in Microsoft deals, on timing, sequencing, and the fiscal year end window. Fourth, the CIO playbook on Microsoft's shift to CSP and NCE licensing, which explains why the annual term question decides whether a pilot is really a pilot. And fifth, Copilot Credits governance at the EA, for the metered rate caps you should be closing alongside the seat rate. Next session closes module four: proving Copilot value. Adoption telemetry, the business case that survives a CFO, and the evidence file that decides the renewal. See you there.