A full worked case: an expiring ULA, a growing OCI commitment, a Fusion renewal, and a hyperscaler alternative, sequenced into one deliberate position. Three knowledge checks along the way, and 1 clip from a senior cloud advisor.
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 thirty of thirty, the capstone. Twenty nine sessions of instruments, meters, clauses, files, and disciplines, and today they all land on one table at the same time, which is exactly how this actually happens in a real estate. The case is deliberately ordinary: an expiring database ULA, an OCI commitment tracking to a shortfall, a Fusion renewal with visible shelfware, a hyperscaler alternative that is genuinely credible, and a support ledger underneath connecting all of it. Four events, twelve months, one vendor. Handled one at a time, that is four sequential defeats. Handled as a package by a team that prepared, it is the single best negotiating opportunity a decade of this relationship will offer. Today: the case, the position that must exist before the first meeting, the sequencing, the package itself, and the endgame. And at the end, the whole course assembled into one sentence. Let's negotiate.
Five takeaways. One, the case: a realistic estate with four events converging inside twelve months, the situation this entire course was built to handle. Two, the position: what must already exist before the first meeting, the numbers, the files, the alternatives, the walk away, the owner, and the uncomfortable fact that if it does not exist by then, the negotiation has already happened without you. Three, the sequence: which event moves first, what feeds what, and why the ordering is worth more than any individual concession you might win. Four, the package: four negotiations turned into one deal, what you ask for, what you concede, and the short list you refuse to trade at any price. And five, the endgame: the last six weeks, the term sheet, their calendar, and the paper watch that protects everything won. Then we close the course. One note before we start: nothing in this session is new. Every move today is a discipline from an earlier session, which is precisely what makes it a capstone rather than a lecture.
The case. Event one, the expiring ULA: your database unlimited license agreement certifies in nine months, deployment grew substantially during the term, and some of that growth sits on AWS, which, per session eleven, counts under the authorized cloud policy. Certification counts are therefore the largest single number in this whole negotiation, and nobody has finished computing them. Event two, the OCI commitment: four million a year, tracking to seventy eight percent burn, expiring in seven months, so you are looking at roughly eight hundred and eighty thousand of forfeit risk plus a resize decision, session seven's exact scenario. Event three, the Fusion renewal: six million across ERP and HCM, uncapped, and your gap ledger shows twenty eight percent of seats inactive across twelve dated quarters, renewing in twelve months. Event four, the hyperscaler alternative: your AWS EDP renews in five months, and Database at AWS is a genuine option for part of the estate, which makes routing, marketplace burn, and real leverage all live at once. And underneath everything, the support ledger, five million a year on the ULA estate, linked to the OCI commit through Support Rewards. Nothing here is exotic. What makes it a capstone is the convergence: four instruments, one vendor, and a twelve month window in which every one can be traded against the others, or surrendered one at a time by a team that never noticed they were connected.
Building the position, and this slide is a checklist of things that must already exist. The numbers: the estate dashboard from session twenty six; certified ULA deployment counts including the AWS estate priced at policy rates; OCI burn trended honestly to expiry; the gap ledger showing twenty eight percent inactive across twelve dated quarters, which is Tom's engineering firm all over again; and the support line netted against rewards accrual. The alternatives: the five row platform comparison from session fifteen, rerun with current numbers; a Database at AWS route priced with marketplace burn against the EDP, session fourteen's triangle; measured numbers from a funded pilot if you took one, session twenty nine; and third party support modeled honestly as the floor that makes the support line negotiable at all. The decisions: a walk away number owned by a named executive who would actually spend it; the target term architecture; the clause ask list, cap, holds, swaps, assignment rights, divestiture adjustment; and one channel with a single owner who chairs everything. Now read that list again and notice what it is not: it is not a strategy invented at the table. It is the standing state from modules five and six, accumulated quietly over the preceding year, which is why a capstone negotiation is far less a performance than a harvest. First check tests whether you will defend it.
First check. With four events converging, your account executive proposes something that sounds excellent: a single grand negotiation covering everything, starting next month. Your ULA certification work is incomplete. The right answer: A, agree, one big deal is exactly the package strategy this course teaches. B, agree to the package in principle, but not to that timing: the ULA certification counts must be finished first, because entering a package negotiation without knowing your own deployment position is bidding against yourself. C, refuse the package and negotiate each event separately. Or D, agree, and use the negotiation itself to discover the certification numbers. Pause here. A package is powerful. What does it require from you that four separate deals do not?
The answer is B, and the skill being tested is separating two judgments that arrive glued together: the package is right, and the timing is a trap. A package works because every instrument becomes tradeable against every other, but it demands something four separate deals never do, complete knowledge of your own position on all four simultaneously, because a package prices as a whole and a weakness anywhere quietly discounts everything. Walk in with unfinished certification and the largest number on the table is one only the vendor can estimate, and session twenty six already told you whose estimate wins when only one party holds data. B keeps the strategy and moves the date, which is almost always the right shape of answer when a good idea arrives with a bad calendar attached. A confuses agreeing to a structure with agreeing to a schedule, and the schedule is precisely the concession being requested, which is why the proposal arrived as a favor. C throws away the course's central insight: four separate negotiations is four separate defeats in sequence, each priced without the leverage of the others, and the ULA in particular is worth far more traded than settled alone. D is the most seductive and the worst, because discovering your own compliance position inside a live negotiation hands the vendor both the finding and its timing, which is session twenty five's claim mechanics volunteered rather than defended. The rule: package the deals, own the calendar, and never let a negotiation start before your arithmetic finishes.
The sequencing decision, twelve months ordered on purpose. Months one to four: certify the ULA properly, taking the time to count the AWS deployment correctly rather than optimistically, rerun the platform comparison, and open the Fusion renewal file at T minus twelve exactly as session twenty two prescribes. That period produces every number the package will need. Month five: settle the AWS EDP, and settle it with marketplace terms that admit Oracle spend, because those terms decide whether the Database at route is even available and whether marketplace purchases can burn the commitment, session fourteen's routing question answered before it matters. Months six and seven: table the package, all four instruments in one conversation, ULA exit, OCI resize, Fusion right size, support treatment. Months eight and nine: close inside Oracle's fourth quarter with the term sheet already agreed, because their Q4 approves clauses their Q1 refuses, session twenty nine's calendar. And months ten to twelve: execute, certification lodged, commit reset, Fusion co termed into the pillar architecture, register updated, which becomes the next cycle's standing state. The sequencing rule from session twenty seven, applied to the hardest case: settle the instrument that shapes the others first, keep the biggest events off one team's desk in one week, and time the close to their calendar rather than yours. Our guest analyst has run exactly this convergence. Tom.
Guest analyst The engagement I would put in a textbook was a European manufacturer, and it looked exactly like today's case: a ULA certifying, an OCI commit heading for a shortfall, a Fusion renewal, and an AWS relationship that mattered. When they called me, four different people were handling the four events, and none of them had met. The first thing we did was not negotiation, it was arithmetic, four months of it. Certification counts finished, including a hard argument about their AWS deployment that we won by getting the counting right first. Burn trended honestly. The gap ledger already existed, thankfully, twenty six percent inactive Fusion seats with dates on them. And a Database at Azure route priced properly through the marketplace. Only then did we put everything on one table, and here is what that changed. Oracle wanted the OCI commit renewed and grown, badly, it was the strategic line in their account plan. We wanted a clean ULA certification, a right sized Fusion base with a cap, and the support line held. So we traded the thing they wanted most against the four things we wanted, in one conversation, rather than letting them win the commit in March and then price the Fusion renewal in September as if the commit had never happened. Final shape: certification accepted at our counts, the commit renewed at a realistic size with the shortfall relieved, Fusion right sized by twenty two percent with a three percent cap and swap rights, and support flat, net of rewards. Their account director told me afterwards it was the toughest negotiation of his year, and then renewed the relationship happily, because the deal was real on both sides. What made it work was not clever tactics. It was that four months earlier, four people who had never met started building one position.
Four people who had never met, four months of arithmetic before a single meeting, and then one table where the thing they wanted most bought the four things the customer wanted. Not clever tactics. One position, built early. Second check is the moment inside that negotiation where packages get dangerous.
Check two, mid package. Oracle offers to waive the OCI shortfall entirely, roughly eight hundred and eighty thousand, if you renew Fusion flat at the current two thousand seats for three years. Your ledger shows fourteen hundred and forty active. What is that offer really worth? A, excellent, an eight hundred eighty thousand shortfall waived for a flat renewal is clearly positive. B, negative: it trades a one time credit problem for three years of five hundred sixty paid unused seats, roughly two and a half million at typical Fusion rates, and the right sizing was the larger prize, so take the shortfall relief and refuse the seat lock. C, neutral, both sides give something up. Or D, excellent, because flat renewals are rare and hard to get. Pause here, and price both halves over three years before you answer. Which number is bigger?
The answer is B, and this is the package's characteristic danger, which is why it gets the capstone's middle check. A genuinely valuable concession is offered against a substantially larger one, and the smaller number is the one wearing the exciting label. So price both legs, on paper, before reacting. The shortfall waiver: a one time benefit near eight hundred eighty thousand, real, legitimate, worth having. The seat lock: five hundred sixty unused seats for three years, which at ordinary Fusion pricing lands somewhere near two and a half million, and worse, it converts your single best piece of evidence, twelve dated quarters showing a twenty eight percent gap, into a contractual base that every future uplift will apply to. The trade is negative by roughly three to one, and it is presented as generosity because the generous half is the visible half. The professional response is not outrage, it is separation: accept the shortfall relief, which they offered and which is genuinely theirs to give, and refuse the seat lock, countering with the right sized quantity, the cap, and price holds for growth. A and D price the label rather than the deal, and D adds a myth worth killing, flat renewals stop being rare the moment a documented right size case exists, because flat is what a vendor offers to avoid a reduction. C mistakes symmetry of gesture for symmetry of value. The capstone rule: in a package, every leg gets priced separately, in writing, before any leg gets accepted.
The package itself, five components. The ULA: certify at the maximum defensible count, including the AWS deployment at policy rates, and then exit rather than renew unless growth genuinely justifies another term, because the certified estate becomes the BYOL asset that every cloud position in this course draws against. The OCI commit: resize from measured burn rather than from anyone's optimism, address the shortfall, and size the new commitment to the certified BYOL estate plus the real pipeline, sessions seven and eight. The Fusion renewal: right size to the ledger, take the cap, the swaps, and the holds, and co term into the target pillar architecture, and note that this is where the clause ask list actually gets spent, so bring it complete. The support line: priced net of Support Rewards from the resized commit, with third party support modeled honestly as the floor that makes the number negotiable at all. And the fifth component, which is the one I would tattoo on a negotiator: what you refuse to trade. The exit capability from session twenty four, the entity and assignment clauses from twenty eight, the down lane, and the right to keep measuring your own usage. Concede money before you concede optionality, because money recurs every year and optionality, once traded away, does not come back at any price.
The endgame, the last six weeks, session twenty one at estate scale. Term sheet first: the entire package on one page, initialed before any order is drafted, quantities, term dates, the cap number, the price holds, the swap rights, the assignment and divestiture clauses, and the support treatment. Structure agrees before price, always, because the moment price is agreed the negotiation is over in the seller's mind and every clause afterwards is bought back at a premium. Close on their calendar: ready in March, willing to wait until May, because their fourth quarter approves clauses their first quarter refuses, and the only buyer who can use that fact is the one who did the preparation nine months earlier. And the paper watch: session seventeen's review run on the final documents, every term sheet line present, every verbal concession typed, definitions read aloud, entity lists checked against the current legal structure per session twenty eight, because deals leak more value in the last seventy two hours than in the preceding nine months, and the leak is almost always something that was won and then simply failed to appear. Then the unglamorous final step in the note, which decides the next cycle entirely: update the register, the calendar, the dashboard, and the files to what was actually signed, and open the next renewal's file that same week. Which is, not coincidentally, the last knowledge check of this course.
Final check of the course. The package closes well: certification accepted, commit resized, Fusion right sized with a cap, support flat. What single action best protects the value you just won? A, announce the savings internally and move the team onto other work. B, update the register, calendar, dashboard, and files to what was signed, and open the next renewal file the same week, because everything won is only preserved by the standing state that produced it. C, lock the contract away and revisit at the next renewal notice. Or D, immediately begin negotiating the next expansion while the relationship is warm. Pause here. Where does the value of a well negotiated package actually leak away over the following three years?
The answer is B, and the reasoning is the quiet thesis of this entire course. Ask where a well negotiated package leaks, and it is never at the table, it is over the following three years: the cap nobody remembered to invoke at renewal, the swap rights that expired unused, the price holds never exercised because the person who won them changed roles, the right sized quantity that drifted back up through mid term additions nobody gated, and the notice window that arrived unowned. Every single one of those is a governance failure rather than a negotiating failure, which is why the correct answer is deliberately unexciting: the register, the calendar, the dashboard, and the files, updated to what was actually signed, are the only mechanism that converts a good day into a good decade. And opening the next renewal file the same week is session twenty two's habit, which makes the following cycle start at T minus thirty six rather than T minus sixty. A celebrates, which is fine, and then disperses the only people who understand the paper, which is not. C is how estates arrive at renewals having genuinely forgotten what they own, the exact condition every trap in this course exploits. D spends fresh relationship capital on new commitments at the precise moment your leverage is lowest, having just signed everything. So the closing rule, and it is the sentence I would leave you with above all others: the deal is an event, the position is a practice, and only the practice compounds.
The course, assembled, and let me spend a moment on what you actually hold now. Modules one and two gave you the paper and the meters: the Cloud Services Agreement, the ordering documents, the policies, the pricing mechanics, and then OCI commercially, credits, commitments, BYOL ratios, Support Rewards, and cost governance. Module three put Oracle on every cloud and handed you the five row method, so no platform question ever has to be answered with a vendor's calculator again. Module four was SaaS: the metric families and their definitions, the order read like an analyst, ERP, HCM, and the migration that moves an entire estate. Module five was the lifecycle, the years between signatures that most estates manage worst: the deal, the renewal, the shelfware, the exit, the compliance machinery. And module six was governance and this capstone: FinOps, contract governance, corporate events, the relationship, and today. Through all of it, three sentences recur, and they are the course. Whoever holds the whole picture runs the relationship. Position, evidence, and alternative get built between events, not at them. And every clause worth having is cheap before it is needed and unbuyable afterwards. If you remember nothing else from thirty sessions, those three will still earn their keep at every table you sit at for the rest of your career.
Session thirty, three sentences, and then we are done. One: four converging events are not four problems, they are one package, provided your own arithmetic is finished before the negotiation opens and every leg of any offered trade is priced separately, because the generous half is always the visible half. Two: sequence deliberately, settle the instrument that shapes the others first, close inside their fiscal pressure with the term sheet already agreed, and concede money before you ever concede optionality. Three: the deal is an event and the position is a practice, so the value is protected by the register, the calendar, the dashboard, and the next file opened the same week, which is the standing state this entire course exists to build. That is Oracle Cloud Management, thirty sessions, complete. Take the module tests to check what stuck, and when you want to prove it, the certification exam is there. Thank you for spending fifteen hours with me. Go and run your estate.
Final homework, about two hours, and it is the only one that matters: build your own capstone. One, map your convergence, every Oracle event in your next eighteen months on a single page, renewals, commitments, ULAs, support anniversaries, hyperscaler dates, and then circle any two that could plausibly be traded against each other, because that circle is your first package. Two, audit the position against today's position slide: which numbers, alternatives, and decisions exist right now, and which are missing, and the gaps are not a criticism, they are your work plan for the next two quarters. Three, draft the sequence: order the events on purpose, one sentence per event explaining what it feeds and why it sits where it sits, and if you cannot write that sentence for an event, it is in the wrong place. Four, write the package on one page: what you ask for across all instruments, what you are willing to concede, and the short list you refuse to trade at any price, and be honest on that last list because it is the one that gets tested. And five, name the owner and the date: who runs it, and what date must preparation start so the close lands inside the vendor's fourth quarter. Put both in the calendar today, before this session gets filed away with the good intentions. That is how a course becomes a capability.
Five final reads, all free on redress compliance dot com, and these five are the shelf I would keep beside any real negotiation. First, Oracle cloud negotiations, the package approach in reference form with worked outcomes. Second, the ULA exit strategy guide, certification and exit, the first instrument in today's case and often the largest number on any table. Third, the Oracle contract clause negotiation playbook, the ask list clause by clause, which is what your term sheet is made of. Fourth, the Support Rewards guide, the arithmetic that links the commit to the support line and quietly decides half of these packages. And fifth, the Oracle top ten negotiation recommendations, the condensed version, genuinely worth rereading the week before any real table. That is session thirty, and that is the course: thirty sessions, six modules, one capability. You came in able to read a quote. You leave able to run an estate: to count correctly, to choose platforms with evidence, to buy subscriptions on definitions you have read, to renew from a position, to leave if you should, and to put four events on one table and negotiate them as one deal. The tests are there when you want to check yourself, and the certification when you want to prove it. Thank you for your time, and good luck at the table. You are ready for it.