The uplift playbook, right sizing with your own usage data, and running year three like a negotiation instead of an invoice. 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 two of thirty. Last week we negotiated the initial deal and I told you the deal you sign is the renewal you run in year three. Well, it is year three now. Today is the renewal session, and it starts with an uncomfortable statistic from the field: most SaaS renewals are not negotiated at all. They are processed. A quote arrives six weeks out, someone winces at the percentage, argues it down a point or two, and signs, and the machine on the other side is calibrated to exactly that behavior. The alternative is not heroics at the deadline, it is a twelve month program that makes the deadline boring: the timeline, the uplift machine and how it prices your silence, the right size case built from your own usage data, the five layer defense stack, and the three traps that convert your renewal into their expansion. The renewal is won at T minus twelve months or lost at T minus thirty days. Let's make sure you are on the right end of that sentence.
Five takeaways. One, the timeline: the twelve month renewal program, built backwards from the notice window, and why the last ninety days should be execution of a settled position, never the formation of one. Two, the uplift machine: where the renewal quote's percentage actually comes from, what the seller's playbook assumes about you, and the compounding arithmetic that makes a settled for four percent feel like a win while costing a third more spend inside a decade. Three, right sizing: your usage data as the counter quote, active against billed, module by module, and the crucial fact that renewal is the one window where quantities travel down. Four, the defense stack: cap, quantity case, benchmarks, competitive check, escalation, five layers in firing order. And five, the traps: the early renewal offer, the repricing threat, and the flat if you add bundle, each one arriving dressed as a favor, each one priced. The session's reframe, worth the whole half hour: the percentage conversation is their home field, the quantity conversation is yours.
The timeline, twelve months, built backwards. T minus twelve: open the renewal file, the usage ledger, the ratios, and the signature memo from the original deal, because evidence takes a year to accumulate and opinions take a meeting, and a renewal argued on opinions loses to a machine argued on defaults. T minus nine: the competitive check, is an alternative real enough to evaluate, and if so, start now, because session twenty one taught you that tension needs months of visible process to become credible, a logo mentioned in week six moves nothing. T minus six: the right size position, target quantities, modules to drop, swap rights to exercise, your ask, defined before their quote exists, because whoever tables first anchors the conversation. T minus four: engage Oracle with your position first, and escalate past the renewal desk immediately, renewal desks are staffed to say the uplift number and very little else, the negotiation lives with the account team and above. And T minus thirty to ninety days, depending on your contract: the notice window, which gets protected in writing before you negotiate a single day past it, because auto renewal converts silence into signature. Now the failure mode this table exists to prevent: starting at T minus sixty. Past the leverage, inside the notice window, negotiating against a quote that anchored first. Everything else today assumes you start at twelve, and the first check is about what to do when you did not.
The uplift machine, three parts. The default: an uncapped contract renews at whatever the current standard uplift is, commonly somewhere in the four to eight percent range depending on the year and the product, applied, and this is the detail that matters, to the full billed quantity, shelfware included. The machine does not ask whether your two thousand seats are used. It assumes the quantity renews unchanged, and every year of your silence trains it to keep assuming. The cap, if you won it: session twenty one's clause now pays its dividend, the percentage argument is over before it starts, and the entire negotiation moves to quantities and modules, which, keep hearing this, is where your leverage actually lives. And the anchor arithmetic, which deserves cold statement: a seven percent ask settled to four feels like a victory, and four percent compounding for a decade is over forty percent more spend, the machine is happy to lose the battle it staged. Here is the counter that changes the game: the answer to their percentage is not a smaller percentage. It is a smaller quantity, priced from your usage file. When the conversation is about seven versus four, you are on their field. When it is about two thousand seats versus fourteen hundred active users, you are on yours, and they know it, which is why the whole renewal playbook on their side is built to keep quantities out of scope. First check.
First check, the late discovery. The renewal quote lands at T minus forty five days: nine percent uplift on all two thousand seats. The auto renewal notice deadline is T minus thirty. Nobody has prepared anything. The first move: A, negotiate the nine down to something reasonable before the deadline. B, serve the non renewal notice in writing before T minus thirty to stop the auto renewal, then negotiate without the clock, while starting next cycle's file at T minus twelve. C, let it auto renew, nine percent is roughly market. Or D, threaten to leave for a competitor. Pause here. What does the notice actually do, what does it cost, and which deadline in this scenario is real?
The answer is B, and it is an exercise in triage: identify the only real deadline and defuse it. The auto renewal notice window is real because auto renewal converts silence into a signed year at the quoted uplift, everything else in the scenario is negotiable afterwards. A protective non renewal notice, served in writing before the deadline, costs nothing and commits you to nothing: you can still renew, on whatever terms you subsequently agree, but the seller's best asset, your clock, is gone. Notice what that does to A: with the deadline defused, negotiating the quote becomes possible instead of desperate, you are no longer bargaining against a calendar that does the seller's work for free. C signs nine percent on two thousand seats without knowing whether two thousand is even the right number, and unprepared, at T minus forty five, nobody knows. D fails every credibility test from last session: an alternative that is unevaluated, unsponsored, unpriced, and announced six weeks before expiry is theater, and renewal desks watch theater all day. Now the honest coda: a renewal discovered at T minus forty five is already lost as a negotiation. The notice preserves what is preservable, a point or two, maybe some flexibility, but the real fix is the last clause of B, the file for the next cycle opens today. Which is what the rest of this session builds.
Right sizing, and the headline is the format: the counter quote is a table, not a percentage. Row one, active versus billed users, from your identity system and the application's own analytics, twelve months of it: this moves the headline quantity, the biggest number in the deal. Row two, module adoption: feature usage per module, because whole modules get dropped or swapped at renewal, and adoption data is how you know which. Row three, workforce actuals: HR headcount against the Hosted Employee counts, session nineteen's metric trued to reality, including the divested division that never left the bill. Row four, the floor ratios from session sixteen: billed against real per line, the floors that were never renegotiated, now up for renegotiation. Row five, benchmark rates for your size and mix, last, because rates come after quantities. Why this works is the asymmetry we have tracked since session sixteen: quantities travel one way mid term, and renewal is the sixty day window when the down lane opens, then closes for three years. Miss it and the shelfware renews, and worse, it renews as precedent. And why the data matters: a right size case built on twelve months of your own measured usage is the one argument a renewal desk cannot wave away as negotiating posture, numbers from their own product's analytics do not argue back. Our guest analyst ran exactly this play, let's hear how it lands.
Guest analyst The renewal I use as the template was a pharmaceutical client, about four million a year of Fusion, ERP and EPM, heading into their fourth renewal. The first three had been processed, not negotiated: quote arrives, controller winces, signs, average uplift just over six percent, compounding quietly. We were engaged fourteen months before the fourth one, which is the single most important fact in the story. Here is what fourteen months buys. Months one through three: instrumentation, twelve hundred billed seats, and we found four hundred and ten that had not authenticated in a year, plus an EPM module bought for a finance transformation that had been cancelled two years earlier, still billing. Months four through nine: the file, monthly usage snapshots, the workforce actuals, benchmark rates from three comparable deals, and a scored evaluation of one workload against a competitor, small, real, and visible. Month ten: we tabled first. Not a response to their quote, our position paper: seven hundred fifty seats, the dead module dropped, swap rights exercised on another, benchmark rate, three year term as the sweetener. The renewal desk said no, repricing, impossible. The account director, at the escalation meeting we had scheduled two months earlier, said, in effect, let me see what I can do. Signed: seven hundred eighty seats, flat rate, the module gone. Annual spend fell from four million to about two point six, against a machine that had opened at four point three. The client asked what the trick was. There was no trick. We just started before the machine did, and brought its own data. The renewal desk processes silence. It has no play for a customer who tables first with fourteen months of receipts.
Four million down to two point six, against an opening quote of four point three, and the whole trick was starting fourteen months early and tabling first with the product's own usage data. The renewal desk processes silence, and it has no play against receipts. Second check.
Check two, the repricing threat. Your usage file shows fourteen hundred monthly active users against two thousand billed seats. Growth plans need maybe a hundred more. Oracle warns that reducing seats triggers repricing: your discount was tied to volume, cut the quantity and the rate gets worse. The right position: A, renew two thousand, the repricing threat makes reduction pointless. B, right size to about fifteen hundred, with the repricing math done in advance: even at a worse rate, five hundred fewer seats usually wins, and term length or a growth price hold can be traded to keep the rate anyway. C, cut to fourteen hundred exactly and refuse any rate discussion. Or D, split the difference at seventeen hundred as goodwill. Pause here. Run both totals: two thousand at the old rate, fifteen hundred at a plausibly worse one. Which is smaller?
The answer is B, and the method inside it matters more than the number: do the repricing arithmetic before the meeting, never during it. The threat is real, discounts are genuinely tiered to volume, but it is also just arithmetic, so run it: two thousand seats at the old rate against fifteen hundred at the threatened rate, and in most real cases the smaller quantity wins comfortably, because discount tiers move rates by points while the shelfware moves the quantity by hundreds of seats. A threat you have already priced is a line item; a threat you have not is a conversation stopper, and the renewal desk knows which one it is dealing with within a minute. B's second half is the professional finish: sellers defend volume because volume is their commission structure, so hand them a different commitment story, a longer term, a growth price hold at the new quantity, the co term consolidation from session twenty one, and the rate usually survives the reduction, they need something to book, it does not have to be your empty seats. A is the threat succeeding exactly as designed: six hundred empty seats renewed to protect a percentage. C cuts to the actuals with zero headroom, no onboarding buffer, and buying seats back mid term prices at list, fifteen hundred is a position, fourteen hundred is a vulnerability. And D, the goodwill midpoint, is session eighteen's lesson wearing renewal clothes: the average of a computed number and an uncomputed one is a donation. Quantities from data, rates from trades, arithmetic before meetings.
The defense stack, five layers, in firing order. Layer one, the cap, if it exists: open with it, the percentage conversation ends and you fight entirely on quantities, which is the fight you want. If it does not exist, calculate out loud what its absence has cost across the term, and make the cap a condition of this signature, renewals are the second and last window where caps get written. Layer two, the quantity case: the right size table, active against billed, modules against adoption, the single largest mover of money in any renewal and the one argument that cannot be dismissed as haggling, because it is the product's own telemetry. Layer three, the benchmarks: market rates for your size and mix, and understand their role, they rarely win alone, their job is to convert the psychology of the room, your number becomes a position, theirs becomes an ask. Layer four, the competitive check: even a partial alternative, one module, one workload, reprices everything, provided it passes the credibility tests, evaluated, sponsored, alive. Tom's client evaluated one workload, and it colored the whole four million. Layer five, escalation: renewal desks quote, account teams negotiate, and the escalation happens on a schedule you set, with your position paper attached, months out, not as a complaint in the final week. Tom's escalation meeting was booked two months before it happened. Five layers, one order: clause, quantity, benchmark, tension, altitude.
The traps, three of them, each arriving dressed as generosity. The early renewal: an attractive rate to renew now, a year before expiry, sometimes with a small discount attached, and here is what it actually purchases, your notice window, your competitive check, your right size analysis, and all of next year's leverage, retired twelve months early. Price it with one question: why would a seller pay to extinguish my options? Because the options are worth more than the payment, always. The repricing threat we just handled: arithmetic, not a wall, run both totals and bring the commitment trade. And the flat if you add: zero uplift if a new module joins the order, and the second check demolishes this one with numbers, so here is just the shape: flat is not flat when the real direction is down, the module rides in on the uplift you supposedly avoided, brings its own shelfware risk, and fattens the base that next cycle's percentage applies to. Any module worth having survives its own business case and its own negotiation; anything that only makes sense bundled against a renewal deadline did not make sense. The common thread across all three: each converts a moment when spend could fall into a commitment for it to rise, wrapped in favor language. And the common antidote is the file, because every trap, without exception, depends on you not knowing your numbers. Last check runs the flat if you add at full scale.
Last check. Oracle offers a flat renewal, zero uplift on your three million dollar base, if you add an analytics module priced at six hundred thousand a year. Your usage file shows thirty percent of existing seats inactive. The analyst's read: A, take it, zero uplift on three million beats a five percent increase. B, decline the bundle: the real move is down, not flat, the right sized base is roughly two point one million, the flat offer costs one and a half million more than that over the year, and the module gets evaluated on its own case, separately, if at all. C, take it but negotiate the module down to four hundred fifty. Or D, counter with flat and half the module. Pause here. What is the actual alternative to the flat offer? Price all three positions: theirs, the naive baseline, and yours.
The answer is B, and the exercise is pricing all three positions, because the offer is engineered so that you only ever price two. Their position: three million base plus six hundred thousand module, three point six million. The naive baseline the offer is built to beat: three million plus five percent, three point one five, and against that, flat plus a shiny module for four fifty more looks nearly reasonable, which is the trap functioning. Your position, the one the offer is designed to keep out of the room: thirty percent inactive seats means a right sized base around two point one million, and against that number, the flat offer is one and a half million a year of extra spend, an expansion wearing a discount's clothes. That is the whole mechanism: the flat anchors the current base as the reference point so the right size case never gets tabled, and the damage compounds, because next cycle's percentage machine applies to three point six million instead of two point one. C and D negotiate the trap's terms instead of refusing its frame, better than A but still inside the wrong conversation, haggling over the price of the anchor. And the fair coda: the analytics module might genuinely be worth six hundred thousand, in which case it survives exactly what B prescribes, its own business case, its own evaluation, its own session twenty one negotiation with clauses and a ramp. The test that never fails: an offer that only makes sense against a deadline is an offer that does not make sense. Renewal desks retire quotas with bundles; your job is to make them retire it somewhere else.
The renewal file, the document that actually runs the renewal, three sections. The record: the original signature memo from session seventeen, what was bought, on which definitions, with which clauses, plus the gap ledger, quarterly, which next session industrializes, plus every mid term addition with its co term stub, because those stubs are where accidental term architecture comes from. The position: the right size table with quantities from twelve months of usage, modules from adoption data, workforce from HR actuals; the repricing math, both totals, pre computed; the benchmarks; and the clause asks, the cap if absent, swap rights, the down lane, each written as a condition of signature. The leverage: the competitive check's status against the credibility tests; the escalation map with names and dates, booked in advance like Tom's; the notice window, diaried at ninety days and protected in writing; and the walk away number, agreed with the executive who would actually have to mean it, because a walk away nobody owns is a bluff wearing a spreadsheet. Opened at signature, fed all year, spent at the table. And the honest question this slide asks of every listener: does this file exist for your biggest renewal right now? If not, the homework starts it, and next session, shelfware and mid term management, is the engine room where its evidence gets built, quarter by quarter, while the account team is busy selling you more.
Session twenty two, three sentences. One: the renewal is won at T minus twelve months, the notice window gets protected in writing before anyone negotiates past it, and a renewal discovered at T minus forty five is preserved, not won, so the fix for this cycle's scramble is opening next cycle's file today. Two: the percentage conversation is the seller's home field and the quantity conversation is yours, so the counter quote is a right size table built from twelve months of your own usage data, with the repricing arithmetic run before the meeting and a commitment trade ready to save the rate. Three: the early renewal, the repricing threat, and the flat if you add all convert a moment when spend could fall into a commitment for it to rise, and every one of them depends on you not knowing your numbers, which is exactly what the renewal file exists to prevent. Tom's version: the renewal desk processes silence, and it has no play against receipts. Next week, the engine room: shelfware, the gap ledger, and mid term management, where the receipts get made. See you there.
Homework, about an hour, and it opens the file this session kept invoking. One, find the date: expiry and notice window for your largest SaaS contract, compute where you are on the timeline, T minus what, and diary the notice date ninety days early, this single calendar entry is the cheapest protection in the entire course. Two, pull the usage: twelve months of active users against billed seats from your identity system, one number per module, the counter quote's first draft, and expect a surprise, there is always a surprise. Three, run the repricing math: current quantity at the current rate against a right sized quantity at a ten percent worse rate, both totals written down, that one page is your armor when the threat arrives. Four, check the clauses: cap, swap rights, down lane, which actually exist on your paper, and whatever is missing goes onto the renewal ask list as a condition of signature, renewals are where missing clauses get their second and final chance. And five, name the owner: who runs this renewal, and does that person know it yet? Because a renewal without an owner at T minus twelve is a renewal the machine wins by default, and the machine, as we established, never sleeps and never starts late. An hour now, or a percentage forever.
Five reads before next session, all free on redress compliance dot com. First, the Fusion SaaS renewal playbook, this session's method in reference depth, Fusion specific, with the timeline expanded. Second, Oracle renewal negotiation strategy, the wider Oracle renewal picture beyond SaaS, because support and cloud renewals rhyme with everything you learned today. Third, the Oracle renewal negotiation checklist, the timeline reformatted as a checklist you can actually run, print it and pin it to the renewal file. Fourth, the enterprise software renewal calendar, notice windows and timing across the whole vendor estate, useful the moment you realize every vendor runs a version of the same machine. And fifth, price holds, caps, and uplift clauses, one more time, because the defense stack's first layer is the clause layer, and it matters that much. That's session twenty two. Start at twelve, table first, bring receipts, and never argue percentages on seats nobody uses. Next week we build the receipts: shelfware and the gap ledger. See you there.