The contract inventory, the renewals calendar as strategy, and co termination across the whole estate. 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 back, session twenty seven of thirty. Last week we governed the numbers, and I left you with a claim: whoever holds the whole picture runs the relationship. Today we govern the paper, because the estate's other picture is contractual, the orders, the commitments, the support sets, the term dates, and it fragments in exactly the same way the numbers did, into folders owned by different teams, filed under different vendors, renewing on dates nobody chose. Here is today's version of last week's claim, and it is the session's title thought: fragmentation is the vendor's architecture, consolidation is yours. Scattered term dates, five clause sets, orders negotiated alone, none of that happened by anyone's plan on your side, and all of it serves the other side's. Today: the contract register including the paper with other vendors' logos on it, the renewals calendar read as strategy, the sequencing board, co termination executed at estate scale, and the cadence that keeps all of it standing. Let's take the paper back.
Five takeaways. One, the inventory: every instrument that governs the estate on one register, and the completeness rule that catches what most registers miss, the non Oracle paper that moves Oracle economics. Two, the calendar: the renewals calendar as a strategy document with three readings, deadlines, collisions, and leverage windows, not the flat list of dates most estates keep. Three, the sequencing: the twenty four month board, which events land when, which feed leverage into which, and the two rules that govern the ordering. Four, co termination: session twenty one's term architecture, which you learned one contract at a time, now executed across the whole estate, pillar events, strays pulled home, and the quiet dividend of clause harmonization. And five, the cadence: quarterly, per event, and annual rhythms that keep the paper governed, so that when any negotiation opens, anywhere in the estate, the position, the paper, and the plan already exist. Preparation as a standing state, not a phase. That sentence is where this module is headed.
The paper estate, five instrument classes on one register. Class one, the CSA and its orders: every order under the master, with its products, metrics, quantities, term dates, the clauses won, and the co term stubs, because the estate's actual SaaS and OCI terms live in the orders, not in anyone's memory of them. Class two, and this is the row that separates real registers from folder listings: the hyperscaler commitments, the AWS EDP, the Azure MACC, the Google CUDs, their balances, their terms, and their marketplace burn rules, on the Oracle register, because session fourteen taught you they absorb Oracle spend and steer Oracle routing, and filing them under another vendor makes the triangle invisible. Class three, the support contracts: every support set, its cost, its coverage, and the repricing linkages between sets, module three's ledger and session twenty's traps, in one place. Class four, the perpetual license base: the BYOL census, owned, allocated, parked, the asset every cloud position draws against. And class five, the adjacent paper: integrator contracts, third party support, anything whose terms constrain or enable the rest, session eighteen's partner sizing entered the estate through a contract that never touched the Oracle folder. The register's test is borrowed from last week: no instrument the account team knows about that you have not listed. Most estates fail on row two, and the first check is about exactly that failure.
The calendar as strategy, three readings of the same dates. Reading one, as deadlines: every term end, notice window, and true up cycle, each with an owner and a T minus twelve trigger, module five's discipline applied estate wide. Necessary, and if you stop here, you have administration, not strategy. Reading two, as collisions: where events stack, two major renewals in one quarter, a renewal landing in the middle of a reorganization, an EDP and a Fusion term expiring on the same exhausted team. Collisions are found on the calendar in advance or discovered in the fatigue in real time, and the desk on the other side, which sees your dates as clearly as you do, knows which negotiation gets the tired team. Reading three, as leverage windows: where events feed each other, an OCI commitment decision landing near a support renewal, with the rewards arithmetic linking them; a hyperscaler commitment renewal settling just before a Database at decision needs its marketplace terms; a migration signature timed to Oracle's fiscal fourth quarter. The calendar shows not just when leverage exists, but where it can be moved to, because, and this is the reframe that makes the whole session work, term dates are negotiable instruments. A date is not when something must be signed. It is when something can be traded, and every date on the calendar is a card. The sequencing board is where the cards get ordered.
First check. A new estate owner builds the contract register properly: all CSA orders, all support contracts, the complete ULA history, current and accurate. And renewals still keep going badly. What is most likely missing from the register? A, nothing, bad renewals have other causes. B, the non Oracle paper that moves Oracle economics: the AWS EDP and Azure MACC balances that marketplace purchases burn down, the CUDs, and the integrator contracts whose sizing assumptions become Oracle quantities. C, Oracle's own internal account plan. Or D, historical invoices going back ten years. Pause here. Session fourteen's triangle: which contracts decided how that deal was routed? And which vendor's name was on them?
The answer is B, and the diagnosis is structural: an Oracle only register is complete about the trees and silent about the forest, because the estate's biggest routing decisions are decided by paper carrying other vendors' names. Walk the evidence from this course. Session fourteen's Database at Azure deal: eleven million a year of Oracle spend routed through the Azure Marketplace, its economics decided by the MACC's burn down rules, an instrument filed under Microsoft. The marketplace private offers, the Support Rewards eligibility on routed spend, the whole triangle: all dependent on commitment balances living in the AWS, Azure, and Google folders. And session eighteen's partner incentive problem entered the estate through an integrator contract that never touched the Oracle drawer, sizing assumptions in a services agreement that became subscription quantities on an Oracle order. Miss these instruments and every cross vendor play this course taught goes dark: you cannot burn an EDP you are not tracking, cannot time a signature against a MACC you have not read, cannot challenge a sizing whose source contract you never listed. C is intelligence you would love to have and never will, the account plan stays on their side of the table. D is history, useful for trend lines, inert as leverage. The completeness rule, worth writing at the top of the register: an instrument belongs on the list if its terms can change what you pay Oracle, whoever's logo is on the cover. Row two is where real registers get built or quietly fail.
The sequencing board, twenty four months on one page, and the worked example on the slide is one shape, not the shape, the discipline is that someone chose the order on purpose. Walk it. Quarter one, the AWS EDP renewal, settled first, deliberately, because its marketplace commitment language determines whether Oracle spend can burn hyperscaler commitments, which arms the Oracle conversations that follow. Quarter two, a deliberate gap: recovery, file building, no major event lands here, and protecting an empty quarter is as much a decision as filling one. Quarter three, the Fusion renewal, whose T minus twelve program has been running all year: the EDP terms are now known, the OCI intent is alive as leverage, and the team is fresh. Quarter four, the OCI commitment resize, with the Fusion outcome known and the sizing taken from measured burn, timed against Oracle's own fiscal pressure. And beyond, the support and BYOL review, after the cloud positions settle, so the repricing interactions get modeled once, on stable ground, instead of three times on shifting sand. Two rules generate every good board. Rule one: no two majors in adjacent quarters for the same team, fatigue is a concession machine. Rule two: every event's output feeds the next event's input, the sequence is a supply chain of leverage. Our guest analyst rebuilt exactly this discipline on an estate that had none, and the numbers are worth hearing. Tom.
Guest analyst The engagement that taught me contract governance at scale was an industrial conglomerate, five subsidiaries, each buying Oracle on its own authority for fifteen years. The archaeology took a month: fourteen active CSA orders, eleven distinct renewal dates scattered across the calendar, five different clause sets, and the fun part, one subsidiary had a five percent renewal cap another subsidiary's team had never heard of, while a third had swap rights the other four lacked. Every order individually defensible. The estate, collectively, indefensible. And the account team was superb, five warm relationships, five sets of quarterly lunches, and, we established later, at least two occasions where a discount conceded to one subsidiary was quietly funded by terms taken from another. Nobody on the customer side could see it, because nobody on the customer side could see, full stop. The rebuild took a year and it was governance, not heroics. One register, every order and both hyperscaler commitments on a single page. A target architecture: two pillar events, ERP family and HCM family, eighteen months apart. Every renewal and every new order thereafter wrote its term date toward those pillars, strays pulled home one touch at a time. And the clause harmonization, which nobody expected to be the headline: the best cap, the best swap rights, and the best exit terms from across the five clause sets became the group template, demanded at every consolidation event, and largely won, because refusing one subsidiary a term another already held is an awkward conversation for a vendor. Net effect by the second pillar cycle: about eighteen percent off the estate's run rate, and, my favorite metric, the account team started asking us what our calendar looked like. The consolidation did not just save money. It moved the architecture from their side of the table to ours.
Fourteen orders, eleven dates, five clause sets, and a cap one subsidiary held that its siblings had never heard of. One register, two pillars, one template, eighteen percent off the run rate, and the vendor asking to see their calendar. The architecture changed sides. Second check.
Check two, the collision. The calendar shows your AWS EDP renewal in March and your Fusion renewal in June, both run by the same three person team, both eight figures. The governance move: A, run both as they fall, dates are dates. B, restructure the spacing deliberately: a short extension or early engagement moves one event, the EDP settles first because its marketplace terms feed the Oracle position, and the team runs one major negotiation at a time, with the first's output as the second's input. C, outsource one renewal entirely and run the other. Or D, co term them to the same month for maximum combined leverage. Pause here. What does the EDP's marketplace language do to the Fusion negotiation's options? And what does a tired team concede?
The answer is B, resting on two facts. Fact one, the dependency: the EDP's marketplace commitment terms decide whether and how Oracle spend can burn down the AWS commitment, session fourteen's routing machinery, so the Fusion team negotiates better armed after the EDP settles, the ordering is not a preference, it is a supply chain, EDP output feeding Fusion input. Fact two, the capacity: session twenty two established that a real renewal is a twelve month program with an intense endgame, and three people cannot run two eight figure endgames a quarter apart, the second one inherits a depleted team, and depleted teams concede clauses, quietly, at the paper stage, exactly where session twenty one told you deals leak. The desks on the other side can read your calendar too; they know which of your negotiations gets the tired team, and they staff accordingly. The move in B is the one buyers systematically forget they possess: term dates are negotiable instruments, a short extension purchased on one contract, early engagement opened on the other, and the collision dissolves for a rounding error against either deal's value. A treats the calendar as weather, something that happens to you. D misapplies a good idea across the vendor boundary: co terming concentrates leverage within one vendor, where spend aggregates on one table, but an AWS renewal brings nothing to a Fusion table except the same exhausted team, cross vendor stacking is pure collision, zero leverage. C fixes capacity and breaks the dependency, the outsourced event's output still must feed the retained event, a coordination seam you paid to create. The rule, and it generalizes: within a vendor, concentrate. Across vendors, space and sequence.
Co termination at estate scale, five disciplines, session twenty one's architecture graduated from one contract to policy. One, pillar events within Oracle: the ERP family on one date, HCM on another, each event carrying enough spend to command a real twelve month program and real executive attention, without the single cliff where everything renews at maximum dependence at once. Two, every new order is a repair: each purchase, expansion, and renewal writes its term date against the target architecture, and strays get pulled home at the next commercial touch, because drift, order by defaulted order, is exactly how the accidental architecture rebuilds itself behind your back. Three, harmonize the clauses, consolidation's quiet dividend and the surprise headline of Tom's story: when orders merge at a pillar event, the best cap, the best swap rights, the best exit terms across the merging paper become the template, and the weakest clause set gets retired, and vendors largely concede this, because refusing one entity a term its sibling already holds is an awkward position to defend. Four, space across vendors: the sequencing board's rule, Oracle pillars, hyperscaler commitments, and other majors deliberately offset, never stacked on one team. And five, mind the stubs: co terming mid term means stub periods billed at full annual rates, session sixteen's warning, so price the stub against the architecture's value, and where possible time the consolidation to renewals, where stubs cost nothing. Architecture by policy, repaired at every touch. That is the whole method.
The vendor management cadence, the paper's operating rhythm, three frequencies. Quarterly: the register reviewed, new orders logged with their clauses the week they sign, term dates checked against the target architecture, the sequencing board rolled forward one quarter, and the QBR run on your agenda per last session. This is an hour, not a program, the register makes it an hour. Per event: every signature updates the register the same week, the clause comparison grid refreshes, and, the habit that separates governed estates from lucky ones, the next event's file opens the day this one closes, session twenty two's T minus twelve trigger automated by ritual. The signature is not the end of the deal; it is the register's input. Annually: the strategy pass, the whole board reviewed against the vendor strategy, the platform comparison rerun from session fifteen, because platform answers drift, commitment postures reset against measured consumption, and the escalation relationships refreshed while nothing is burning, because the worst time to meet an account director's boss is the week you need them. The cadence's purpose is the same as the dashboard's was: when any negotiation opens, anywhere in the estate, the position, the paper, and the plan already exist. Preparation is not a phase that precedes events. It is the standing state between them. Last check, and it is the org chart's turn.
Last check. A group runs decentralized purchasing: five divisions, each negotiating its own Oracle orders on its own authority. The account team is excellent, and maintains warm relationships with all five. Predict the estate's paper, and name the fix. A, five well negotiated contracts, decentralization brings market discipline. B, scattered term dates, five different clause sets with none complete, quantities that ignore group leverage, and a vendor who arbitrages the divisions against each other; the fix is central contract governance, one register, one architecture, one channel, even if budgets stay divisional. C, identical outcomes to central purchasing, with more paperwork. Or D, a problem only if the divisions are in different countries. Pause here. Who is the only party that sees all five negotiations? And where has this course seen that asymmetry before?
The answer is B, and you have already met this structure twice: it is session twenty six's asymmetry, the vendor holding the whole picture, rebuilt at the level of the org chart, and it is Tom's conglomerate before the rebuild. Trace what the structure produces. Five divisions negotiating fractions of the group's spend, on scattered dates, generate the accidental architecture automatically. Five separate clause negotiations produce five incomplete sets, a cap here, swap rights there, none holding the full session twenty one stack, and no division knowing what its siblings won. Quantities price at divisional volume while the group's aggregate leverage sits unused, nobody's job to spend it. And the sharpest edge, the arbitrage: the vendor, at all five tables, quoting division C a rate division A fought for, only after division C concedes a term division A kept, value moved quietly between negotiations that never knew they were connected, conducted, note, through excellent relationships, which is why the warmth in the question is not the good news it sounds like, five warm channels is five leaks in the message discipline. The fix is deliberately surgical, and it is the difference that makes B achievable in real organizations: governance centralizes, budgets do not have to. Divisions keep spending authority, while the register, the target architecture, the clause template, and the single channel become group functions. A mistakes fragmentation for a market: five buyers of one vendor are not competition, they are a divided customer. C ignores that the vendor prices the division it faces, not the group it cannot see. D imagines geography is the variable; the arbitrage works fine in one postcode. The session's rule, one last time: fragmentation is the vendor's architecture, consolidation is yours, and the architecture that stands is the one somebody designed on purpose.
The governance file, module six at its halfway mark. From last session, the numbers layer: the estate dashboard with its four panels, the operating rhythm, the forecast that produces your number before their quote, chargeback aimed at cost drivers, and the QBR owned. From today, the paper layer: the contract register with the cross vendor rows that make the triangle visible, the calendar read three ways, the sequencing board with its two rules, the co termination policy with clause harmonization, and the cadence that keeps the register alive. Two layers, one test, and I will keep repeating it because it is the module's thesis: when a negotiation opens anywhere in the estate, the position, the paper, and the plan already exist. What remains. Session twenty eight, corporate events: M&A, divestiture, and reorganization under cloud agreements, what transfers, what does not, and why the entity clauses you have been noting since session sixteen become the whole game the day the org chart changes. Session twenty nine, the relationship itself: account team dynamics in a consumption world, quarter ends, funded pilots, and OCI intent as leverage everywhere. And session thirty, the capstone: an expiring ULA, a growing OCI commitment, a Fusion renewal, and a hyperscaler alternative, on one table, sequenced into one deliberate position, the entire course as a single negotiation. The standing state is nearly built. Three sessions remain.
Session twenty seven, three sentences. One: the contract register lists every instrument whose terms can change what you pay Oracle, whoever's logo is on the cover, and most registers fail first on the hyperscaler commitments, the EDP and MACC balances filed under other vendors while quietly steering Oracle routing. Two: the calendar becomes strategy when its dates are read three ways, deadlines, collisions, and leverage windows, and the sequencing board orders events by two rules, no majors stacked on one team, and every event's output feeding the next event's input, because term dates are negotiable instruments, not weather. Three: within a vendor, concentrate, pillar events with harmonized clauses, the best terms across the estate becoming the template; across vendors, space and sequence; and where purchasing is fragmented, centralize the governance even if the budgets stay divisional, because five tables against one vendor is arbitrage waiting to run. Next week: corporate events, the day the org chart changes and the entity clauses wake up. See you there.
Homework, about an hour, the register and the board. One, list the instruments: every CSA order, support contract, hyperscaler commitment, and integrator contract that touches Oracle economics, one row each, term, key clauses, owner, and resist the urge to perfect it, a rough complete register beats a polished partial one. Two, find the seams: mark the rows currently filed under other vendors, the EDP, the MACC, the CUDs, and if they were not on your Oracle list before tonight, congratulations, you have found the gap this session exists to close. Three, draw the board: the next twenty four months of events on one page, circle the collisions, arrow the dependencies, and write down which order the majors should actually run in, then compare that to the order they are currently scheduled to run in, the delta is your restructuring agenda. Four, grade the architecture: your Oracle term dates against the pillar model, co termed, staggered, hybrid, or accidental, name it honestly, and note the next order or renewal where a repair is free. And five, compare the clauses: cap, holds, swaps, exit terms, across every order, on one grid, because the best of each is your new template and the gaps are the next event's ask list, Tom's five percent cap that nobody else had heard of is sitting in your estate somewhere too. An hour of paperwork, and the architecture starts changing sides.
Five reads before next session, all free on redress compliance dot com. First, a guide to Oracle vendor management, the cadence and relationship structure in reference form, today's third act expanded. Second, FinOps and enterprise software governance, the numbers layer and the paper layer joined into one operating discipline, sessions twenty six and twenty seven in a single frame. Third, the enterprise software renewal calendar, the calendar discipline across the whole vendor portfolio, because the sequencing board works best when every vendor's dates are on it. Fourth, Oracle MOSA versus MCA contract vehicles, the actual vehicles your register tracks and how their structures differ, useful when the register turns up paper from different eras. And fifth, the software renewal management guide, the per event discipline that keeps the register alive between annual passes. That's session twenty seven. The register with the seams found, the calendar read three ways, the board with its two rules, and the clauses harmonized upward. Fragmentation is their architecture, consolidation is yours, and yours is now designed on purpose. Next week, the org chart moves: M&A, divestiture, and the clauses that wake up. See you there.