A ULA term has five endings, and the renewal campaign mentions two. This session closes module 3 with the decision made properly: certification as the default that needs nobody's permission, renewals priced as brand new entry negotiations against a fresh ceiling, the restructure and hybrid paths the campaign never offers, the perpetual ULA tested against futures you do not expect, and the final twelve months run so the decision lands internally before the pressure season starts.
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.
A taught session with three knowledge checks: which path is the default (certification, and everything else carries the burden of proof), the PULA offered at 1.8 times the renewal fee, and the $3.4M renewal against a $1.25M forecast ceiling. It closes with three companies choosing three different right endings by the same method.
The full narration of this session, section by section, for reading and reference.
Welcome back, session fifteen of forty, the last session of module three. Sessions eleven through fourteen built the whole ULA machine: the economics, the entry, the term, the exit. One question remains, and it's the question the entire module funnels into: how should the term actually end? Here's the trap hiding inside that question. By the final year, Oracle's renewal campaign is running, the options presented are two, renew comfortably or certify dangerously, and most companies pick from that menu. But the real decision space has five paths, three of which the campaign never mentions, and the right one is found with arithmetic you can do months before anyone calls. Today: the five paths, the framework that matches your estate to the honest one, the renewal priced as what it really is, the perpetual ULA examined without romance, and the last twelve months run so deliberately that the decision is made before the campaign even starts. Module three ends today, on your terms. Let's decide.
Five takeaways. One, you'll map the paths, all five ways a ULA term can end, not the two the renewal conversation offers. Knowing the full menu changes the negotiation before it begins. Two, you'll apply the framework: growth outlook on one axis, record quality on the other, and the honest path falls out of the intersection. Three, you'll price a renewal properly, as a brand new entry negotiation with a brand new ceiling, never as a rollover of last term's fee plus uplift, which is how renewals are anchored and how they overcharge. Four, you'll judge the PULA, the perpetual ULA, on its actual merits: rarely right, completely irreversible, and occasionally, for a very specific kind of estate, genuinely the answer. And five, you'll run the last year, the final twelve months as a project with milestones, where the decision lands internally at T minus nine, before the pressure season is even scheduled. One dependency to name up front: everything today assumes the count can be done. If session thirteen's records exist, all five paths are open. If they don't, the menu shrinks to one, and it's the expensive one.
Four numbers. Five, the real paths from a ULA term end. The renewal call presents one, occasionally two. The gap between the presented menu and the actual menu is where most of the money in this session lives. Twelve, the months of runway in which this decision gets made deliberately, or gets made for you by default, fatigue, and fiscal quarter. Zero, the number of good decisions made from fear at T minus one. You saw the extension trap last session; every term end trap works the same way, and every one of them dissolves when the count is already done. And twenty two percent, the annuity, which shapes all five paths differently: certification freezes it at the old fee, renewal re bases it on a new one, the PULA makes it permanent. How the term ends is also a decision about what the annuity compounds from for the next decade, which, not coincidentally, is the perfect bridge to module four, where that annuity becomes the entire subject. Sessions eleven to fourteen built the machinery. Today the machinery earns its keep.
The five paths. Path one, certify and exit, the default. Deployments convert to perpetual entitlement, support stays flat at the old fee's annuity, and future growth buys à la carte at benchmark discounts. Requires nobody's agreement, costs no new money. Path two, renew: another unlimited term at a new fee. Genuinely right when growth in the listed products honestly continues at scale, and priced correctly only when treated as a brand new entry negotiation, we'll do that properly in a few slides. Path three, restructure: a new ULA with a different shape, trimmed list, different products, narrower scope, for estates whose growth moved sideways. The campaign rarely mentions this one because it usually means a smaller fee. Path four, the PULA, perpetual unlimited: no certification ever, larger fee, permanent annuity, completely irreversible. It gets its own knowledge check. And path five, the quiet winner in more estates than you'd guess: certify plus buy. Certify everything, take the perpetual position, then make one targeted à la carte purchase for the single area still growing. Two right sized moves instead of one wrong sized one. Note the asymmetry: four paths need Oracle's agreement. The first needs only your records, which is exactly why it anchors the pricing of all the others. The framework, next.
The decision framework, five situations, five honest answers. Growth flattened and records solid: the forecast lands well below any plausible fee, so certify and exit, the default doing its job. Growth strong in the listed products, and here the bar matters: the forecast clearly beats the fee plus its decade of annuity, then renew, negotiated as a fresh entry with everything reopened. Growth moved to other products, the database stabilized but analytics is exploding: the current list is simply wrong, so restructure around where growth actually went. A permanent, stable Oracle core with a decades long horizon and no exit ever desired: the PULA conversation is legitimate, priced across decades, not budget years. And growth modest and concentrated, one area expanding while the rest coasts: certify everything, then buy that one area à la carte, the hybrid. Two inputs power this whole table, and both must be honest: the forecast, built session eleven's way from funded plans, and the record quality, measured session thirteen's way with a dry run. Feed the framework wishful growth or imaginary records and it outputs the renewal machine's answer, wearing your handwriting. Which raises the question of where analysis should start. Knowledge check one.
Knowledge check one. Term end approaches, and the paths are being weighed. Which path is the default, the one that needs no justification? A, renewal, continuity is the safe assumption. B, certification: it converts what is already paid for, and every other path must beat it on arithmetic. C, whichever path Oracle recommends for your situation. Or D, the PULA, because it ends the question permanently. Pause here, and ask: which path requires nobody's agreement?
The answer is B. Certification is the only ending that requires no new money, no new negotiation, and no consent from anyone: the deployments convert because the contract says they convert, the support stays flat, and the estate walks out owning its position. That makes it the analytical baseline, and it places the burden of proof exactly where it belongs: any path that costs new money, renewal, restructure, PULA, must beat certification plus à la carte growth across a decade of honest numbers, or it loses. This framing is worth more than any single negotiation tactic in the module. A is precisely how the renewal machine wins: continuity feels safe, requires no meetings, and quietly costs seven figures; safe feeling and cheap are different properties. C outsources a multi million dollar decision to the counterparty, whose recommendation across thousands of these conversations has been, let's say, consistent. D confuses finality with correctness: ending the question forever is only a virtue if the answer was right, and irreversible answers deserve the most scrutiny on the menu, not a pass. So: price everything against certification, and make the expensive paths argue. Sometimes they win the argument. They never get to skip it.
Pricing a renewal honestly, four disciplines. First, new forecast, new ceiling. Session eleven's method runs again from scratch: honest growth for the coming term, priced à la carte at benchmark discounts. And notice something powerful: the ceiling is lower this time, because the certified position you'd hold on exit covers everything already deployed, only genuinely new growth needs paying for. Renewals must clear a higher bar than original ULAs, not a lower one. Second, the whole session twelve checklist reopens. The product list, the definitions, cloud counting, certification mechanics, caps: a renewal is an entry, and entry is when paper improves. A renewal signed on the old paper is a wasted leverage moment of the largest size. Third, watch the fee anchor. Renewal proposals anchor on the old fee plus a healthy increase, because the old fee is psychologically available. Reject the anchor entirely: last term's fee measured last term's growth expectations, and the only numbers that price this deal are the new forecast and the new ceiling. And fourth, certify first when the timing allows. A completed certification converts your walk away from a threat into a fact, you negotiate the new term while owning the perpetual position, not instead of it. Renewal from strength is a purchase. Renewal from fog is a ransom. Now, the irreversible option. Knowledge check two.
Knowledge check two. Late in renewal talks, Oracle proposes converting to a perpetual ULA at one point eight times the renewal fee: unlimited forever, no certification ever. When is that genuinely attractive? A, almost always, never counting again is worth a premium. B, never, perpetual agreements are always traps. C, narrowly: a permanent, stable Oracle core, no exit ever desired, and the premium beats decades of alternatives, priced honestly. Or D, whenever the immediate budget can absorb the fee. Pause here. What does irreversible require of the analysis?
The answer is C, and the reasoning generalizes to every irreversible decision you'll ever price. The PULA removes the exit permanently, in both directions: no certification ever required, and no certification ever available. The unlimited right becomes permanent; so does the annuity, and so does the relationship. For the rare estate that genuinely is Oracle to the bone, forever, at growing scale, that trade can win, and pricing it honestly means comparing the premium plus its permanent annuity against decades of alternatives, explicitly including the futures you don't expect: the estate that shrinks, the migration nobody's planned yet, the acquisition that changes everything. Irreversible decisions must be tested against surprise, because surprise is the one thing they can't adapt to. A buys the comfort of never counting and forgets that never counting also means never leaving through the front door. B overcorrects into superstition; defensible PULAs exist, the reading list covers their traps precisely because real companies sign them. D is budget thinking, session nine's retired error: affordable and correct remain different questions at every size. And the procedural note matters as much as the analysis: PULA proposals tend to arrive late in renewal talks, priced against decision fatigue, sometimes with a deadline. An irreversible proposal on a deadline answers itself: the deadline is evidence against it. Now, the reshaping options.
The restructure, four moves, for the estate whose growth story changed shape. Move one, trim the list. Products whose growth finished come off the successor agreement, and the fee falls with them. A renewal carrying last term's full list is usually paying unlimited prices for growth that already happened and got certified. Move two, swap the growth. If the expansion moved, middleware done, analytics beginning, the new list follows the new story, priced on its own forecast. The successor is a different deal because the estate is a different estate. Move three, resize the scope: a division level agreement instead of enterprise wide, a two year term instead of three, a tighter customer definition. Unlimited is a feature that can be fitted to the actual growth, rather than draped over the whole company at full price. And move four, always on the table, the hybrid: certify the finished growth into perpetual licenses, and cover only the genuinely growing area with new paper, a narrow ULA or a plain à la carte purchase. Two small right sized deals nearly always beat one large wrong sized one, and the only reason estates don't take this path more often is that nobody across the table suggests it. It's your job to suggest it. Now, the calendar that makes any of this possible.
The last twelve months, run deliberately, five milestones. T minus twelve: price the paths, all five, with real numbers, forecast, ceiling, fee anchors, decade totals, one page per path. This is a week of work that governs millions, and it happens before any conversation with Oracle. T minus nine: align inside. Finance, IT, and the executive who'll sign agree the ranking and the walk away, internally, in a meeting Oracle doesn't attend. This is the single most important date on the slide: the decision lands here, before the renewal campaign is even designed. T minus six: negotiate the chosen path, if it needs negotiating. Renewal or restructure talks open now, and, crucially, the certification runway from session fourteen runs in parallel, visibly, because the count proceeding is what keeps every proposed number honest. T minus three: execute the count. Whatever the path. If certifying, it's the deliverable; if renewing, it's the leverage; either way it happens, on schedule, per the clause. And T zero: land it. Certify on schedule, or sign the negotiated successor with the count done and the improved paper attached. On your date, on your paper, per your decision from T minus nine. The campaign, when it arrives, finds the decision already made. That's the entire trick, and it costs nothing but a calendar.
Term end negotiation dynamics, five rules. Rule one, the count is the leverage. A completed, defensible count makes certification a real, executable walk away, and a real walk away prices every other path down. Fog does the opposite: it prices every path up, because the alternative to whatever Oracle proposes is an exit you can't confidently execute. Rule two, and this is the operational core of the session: never negotiate uncounted. Renewal talks entered before the count is finished are conducted entirely in the currency of fear. Finish the count, then talk. Always, without exception, in that order. Rule three, their calendar still applies. Session nine's fiscal quarters govern renewal deals like any other deal: a term ending mid quarter can often land its successor at the quarter's close, with the discounting authority that date unlocks. Your term end date and their fiscal calendar are two different clocks; use both. Rule four, bundle the entry terms. Any successor agreement reopens every session twelve term, cloud counting, inclusion language, audit terms, caps, and the successor should be strictly better paper than the original, because you know more now, and because the signature is the leverage. And rule five, let certification finish the talks when the numbers say exit. The declaration ends the negotiation politely and completely, and the standing ability to do that is what made every other number honest all along. Final check, with real numbers.
Knowledge check three. The forecast says forty more processors over the next three years, about one and a quarter million dollars à la carte. The records are solid, the count is ready. Oracle offers renewal at three point four million. What does the framework say? A, renew, three point four million for unlimited beats any per unit purchase. B, certify, then buy the growth à la carte: the renewal costs nearly three times the honest alternative. C, counter at two point eight million and split the difference. Or D, take a PULA instead, to avoid deciding again in three years. Pause here. What's the ceiling, and what beats it?
The answer is B, and the arithmetic takes thirty seconds. Certification converts the current estate at zero new cost, support flat. The growth, forty processors, buys à la carte for about one and a quarter million at benchmark discounts, adding roughly two hundred seventy five thousand a year of support. The renewal costs three point four million up front, plus twenty two percent of that, about seven hundred fifty thousand a year, indefinitely, for the same outcome plus flexibility the forecast just said won't be used. The renewal loses by more than two million dollars immediately and roughly half a million a year forever after. No amount of unlimited romance in option A survives contact with that comparison. C is the anchoring trap in its purest form: countering three point four legitimizes a number the ceiling already refuted, and splitting the difference on a refuted number is paying half a mistake. If a counter exists at all, it starts below one and a quarter million, at which point Oracle usually stops countering, which is itself the answer arriving. D spends even more money to avoid a decision the framework just made for free, twice over. And notice, one last time, what made B available: records. With fog, the count is frightening, the walk away is theoretical, and three point four million starts sounding like safety. The entire difference between this company and the one that renews from fear was built in sessions thirteen and fourteen, quarter by boring quarter. That's the module's real lesson, and it compounds better than the annuity.
Three companies, three different right answers, same method. The stabilized estate: growth done, two hundred ten processors deployed, records solid from three years of session thirteen discipline. It certified, exited, kept support flat, and buys its occasional growth à la carte. The default path, executed by the book, and the cheapest ending available. The data platform: listed products genuinely doubling again, a forecast that clearly beats any plausible fee across the decade. It renewed, at two point six million, negotiated as a fresh entry at maximum leverage, with cloud counting, automatic inclusion, and a capped annuity in the successor paper. The expensive path, chosen with open eyes, at a defended price, on better terms than the original. And the shifting estate: database flat, analytics exploding, cloud native platforms on the horizon. It certified the stable base into perpetual licenses, then signed a narrow new ULA covering only the analytics growth, a smaller fee, a precise list, a shorter term. The hybrid, engineered rather than accepted. Three estates, three endings, zero reflexes: each one priced five paths at T minus twelve, decided internally at T minus nine, and negotiated, or didn't need to, from a finished count. The framework doesn't have a favorite path. It has a favorite process. That's the session, and that's the module.
Session fifteen in three sentences, and with it, module three. One, a ULA term has five endings, certification is the default that needs nobody's permission, and every path costing new money carries the burden of beating it across a decade of honest arithmetic. Two, renewals and restructures are new entry negotiations, priced from a fresh forecast against a fresh ceiling with every entry term reopened, never rollovers of last term's fee, and the PULA, being irreversible, gets tested against the futures you don't expect. Three, the decision lands internally at T minus nine with the count already running, because a finished count is the leverage inside every path and the deliverable of the default one. Module three complete: the economics, the entry, the term, the exit, and the decision. Next session opens module four, and it's been foreshadowed all course long: the economics of Oracle support. The twenty two percent annuity, the uplift that compounds it, the policies that stop it ever going down, and the margin that explains why. You've been paying this bill in every worked example since session one. Time to take it apart. See you in session sixteen.
Homework, about an hour, and it's the decision rehearsed. One, price the five paths for your ULA, real or hypothetical: one page each, decade totals, honest forecast. The exercise is the framework, and doing it once on paper makes doing it for real routine. Two, find your T minus nine: working back from the actual term end date, when must the internal decision land? Put the meeting in the calendar now, years early if necessary; calendar entries survive reorganizations better than intentions. Three, draft the renewal ceiling: next term's growth, à la carte, at benchmark discounts. That single number is where every renewal conversation should start, and whoever brings it anchors the room, session twelve's lesson, recurring. Four, stress test the PULA: write down the three plausible futures that would make a perpetual commitment wrong for your estate, a migration, a divestiture, a platform shift. If you genuinely can't find three, the PULA conversation might be legitimate; if you can, you've just written your answer to it. And five, pull a support invoice, your largest Oracle support renewal, and actually look at it. Module four starts next session, and that document is about to become the most interesting piece of paper in the building. That's the hour, and that's module three. See you in session sixteen.
Five reads, all free on redress compliance dot com. First, Oracle ULA renewal negotiation tactics, the renewal path negotiated properly, in full depth, today's session eight expanded. Second, negotiating and managing an Oracle PULA, the ten contract traps in the perpetual variant, mandatory before any perpetual conversation gets serious. Third, the Oracle ULA negotiation playbook, the module's standing reference, decision chapter included, keep it with the agreement. Fourth, the case study, a Fortune 500 retailer's ULA, a real term end decision with real numbers, worth reading against today's framework to watch it work in the wild. And fifth, the Oracle renewal negotiation checklist, which bridges directly into module four: renewals as an annual discipline rather than an annual surrender. That's session fifteen, and that's module three: five sessions, one contract, and a decision framework that prices every ending against the one you already own. The ULA is done. The annuity it leaves behind is not, and that's where we're going. Module four, the economics of Oracle support, begins next session. See you there.