A ULA has exactly one real negotiation, the weeks before signature, and after it the terms are fixed for three years. This working session marks up a proposed $4.2M deal: the product list scoped from nine products to four, the acquisition clause and worldwide territory written in, cloud counting added to the certification clause while it costs a sentence, and the fee anchored to the buyer's own forecast instead of the word unlimited.
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 working session with three knowledge checks: the whole-stack-for-15-percent counter offer priced over a decade, the mid term acquisition that the customer definition decides, and the $4.2M fee against a $3.2M forecast ceiling. It closes with the entry term sheet, Oracle's draft beside the negotiated version, row by row.
The full narration of this session, section by section, for reading and reference.
Welcome back, session twelve of forty. Last session you learned what a ULA is, and the sentence that matters most from it was this: the boundaries are the contract. Today we draw the boundaries, because a ULA has exactly one real negotiation, and it happens in the weeks before signature. After that, the terms are welded shut for three years, and when they reopen at exit, the leverage belongs to whoever wrote the entry paper better. So this is a negotiation session, in the spirit of session ten's working session: there's a proposed deal on the table, a four point two million dollar fee, a draft product list with nine products on it, and a certification clause that never mentions the cloud. By the end of today you'll know how to scope the list to real growth, size the definitions for an unlimited grant, write the exit mechanics while they cost a sentence, and anchor the fee to your own forecast instead of to the word unlimited. Three knowledge checks, and a before and after term sheet at the end. The window is open this week. Let's use it.
Five takeaways. One, you'll scope the product list, the single highest value decision on the paper, in both directions: too broad and you're buying shelf that bills support forever, too narrow and you're manufacturing session eleven's off list findings. Two, you'll write the definitions, customer, territory, and the M&A language, sized for a grant where every boundary is worth ten times more than in an ordinary order. Three, you'll write the exit at entry: certification mechanics, cloud counting first among them, agreed now, while agreement is cheap. Four, you'll price the unlimited period honestly, and I mean a specific arithmetic, not a sentiment: your forecast, the à la carte ceiling, and the decade of annuity that rides on whatever fee gets signed. And five, you'll run the negotiation itself, because a seven figure signature is one of session eight's leverage moments at maximum size, and the clauses that were hard rungs on the ladder in session ten become reachable here. One session, one markup, three years of consequences. Here's why the timing matters so much.
Four numbers on the premise. Four point two million dollars, the proposed fee on today's example deal. Whether that number is a bargain or a donation is genuinely undetermined right now, it depends entirely on work done this session, which is the whole point. Ten, the entry terms on today's checklist, and every one of them follows the same rule: a sentence now, or a fortune later. The cloud counting sentence is the famous example, but the acquisition clause and the support cap run it close. Three years, until any of this can be revisited. Module two taught the timing rule, fix clauses at leverage moments because between them nothing moves. A ULA is that rule with the window welded shut: after signature there are no renewals, no anniversaries, no purchase events, just the term, running. And twenty two percent, of whatever fee gets signed, every year, indefinitely. Hold that one close, because it reframes the entire fee conversation: you are not negotiating a price, you are negotiating a perpetuity. A million dollars off the fee is a million now plus two hundred twenty thousand a year forever. That's the stakes. Here's the map.
The entry checklist, five blocks, and you'll recognize the shape, it's session ten's five block reading, applied to the biggest order form Oracle prints. Block one, the product list: which programs get the unlimited grant. We'll spend a whole section there, because it's the highest value block. Block two, the definitions: customer, territory, affiliates, who may deploy, where, and what happens when the org chart moves mid term. Block three, and this is the block that distinguishes ULA negotiation from everything in module two: certification mechanics. How the exit count will work, written now, method, scope, cloud treatment, timeline, dispute handling. Oracle's draft will be strikingly quiet here, and the quietness is not an accident, vague certification language is worth a renewal. Block four, the commercial terms: the fee, the support base it creates, caps on the annuity's growth, and what happens to your existing support streams, the fold in question, which we'll flag when we price. And block five, the session eight block, because a ULA is still an Oracle agreement: audit terms, divestiture rights, price holds on products that stay off the list. The standard markup rides along. One observation ties the five together: Oracle's draft is precise where it bills you and vague where it protects you. Session ten's finding, at triple the stakes. The markup does the opposite. Block one first.
The product list, Oracle's draft against the honest call, line by line. Database Enterprise Edition: keep, obviously. It's the growth engine, it's why a ULA is on the table at all. RAC, Partitioning, and the packs: keep. They scale with every EE deployment; leaving them off while EE grows would manufacture off list findings by design. WebLogic Suite: question mark, and treat the question honestly. Is there real middleware growth in the plans, or is this line padding? If the forecast says growth, it stays and gets priced on that forecast. If not, it goes. Analytics and GoldenGate: remove. No growth plans, no deployments coming, and every product on the list raises the fee, and the fee sets the annuity, forever. These two lines are pure fee inflation. And Advanced Security, which you may remember from session ten's deal sheet, the stowaway, back again, nobody asked for it this time either. Remove. The scoping principle in one breath: every product on the list you don't deploy is shelf, priced into a perpetuity, and every product off the list you do deploy is an audit finding. So the list is scoped to the growth, precisely, and nothing else. Which sets up Oracle's favorite counter offer. Knowledge check one.
Knowledge check one. Oracle counters your trimmed list generously: add the full technology stack, everything, to the unlimited grant, for only fifteen percent more on the fee. Generous? A, yes, maximum coverage for a modest premium removes all compliance risk. B, yes, because a broader list always certifies more licenses at exit. C, not usually: the premium buys mostly shelf, raises the annuity forever, and the risk it removes is better handled by scoping discipline. Or D, no, because adding products to a ULA list is not allowed. Pause here, and price what fifteen percent more actually costs over a decade.
The answer is C, and the arithmetic is worth doing out loud. Fifteen percent on a four point two million dollar fee is six hundred thirty thousand dollars, now. Plus twenty two percent of that, about a hundred thirty nine thousand, every year, indefinitely. Call it two million dollars over a decade, for unlimited rights to products with no deployment plans, which will therefore certify as nothing at exit, because certification counts deployments, not list entries. Two million dollars for shelf, wearing an unlimited costume. It's session nine's inflated basket, scaled up. Now the wrong answers, because each one is a real sales motion. A sells insurance, and compliance peace of mind is genuinely worth something, but the risk it insures, off list deployment, is handled nearly free by scoping discipline and a provisioning gate, last session's one page on one wall. Buying two million dollars of insurance against a risk you can fence out for nothing is not caution, it's margin. B has certification exactly backwards, you certify what you deploy, and no list entry ever deployed itself. D is simply wrong, lists are negotiated freely at entry, that's the whole session. One honest footnote: the fair version of this trade exists. If middleware genuinely might scale, put WebLogic on the list, priced against its own forecast. One product, one forecast, one price. The whole stack for fifteen percent is not that conversation. Definitions next.
The definitions, four of them, sized for an unlimited grant. First, the customer definition: all majority owned affiliates, current and future. In an ordinary order this clause decides coverage; in a ULA it decides whose deployments legally exist, during the term and, critically, in the exit count. A subsidiary outside the definition deploying listed products isn't enjoying the unlimited grant, it's accumulating findings. Second, the acquisition clause, and this one is ULA specific: companies acquired mid term, do their deployments join the grant and count at certification? Written at entry, this is routine, a sentence with automatic inclusion language. Discovered missing at exit, it's a dispute, conducted at Oracle's leverage, about deployments that already happened. We'll test it in a moment. Third, territory: worldwide, explicitly. An unlimited grant with silent country limits is a compliance program with extra steps, and the moment workloads sit in cloud regions, silent limits get worse, because nobody chose where the failover went. And fourth, divestiture rights, session eight's transition clause at ULA scale: when a division is sold, its deployments need somewhere to go, and the answer, six months of transition rights, say, should predate the transaction by years. All four cost sentences now. Let's test the expensive one.
Knowledge check two. Eighteen months into the ULA, your company acquires a firm running thirty processors of listed products. Do those deployments join the unlimited grant and, at exit, the certification count? A, yes, automatically, the ULA covers the whole company. B, only what the ULA's customer definition and acquisition language say: without it, the acquired estate sits outside the grant. C, yes, provided the acquired company retires its own Oracle contracts. Or D, no, acquisitions always require a new ULA. Pause here. It's session eight's question, at unlimited scale.
The answer is B. The grant covers the entities in the definition, and the definition was written at entry, eighteen months before anyone knew this acquisition existed. That's the entire discipline of definitions: they're written for futures nobody can name yet. With automatic inclusion language, the story is beautiful: the acquired firm closes, its thirty processors join the grant that afternoon, deploy and grow freely for the rest of the term, and fatten the certification count at exit, permanent entitlement, acquired for a sentence. Without the language, the acquired estate is a separate license position that must be tracked separately, under its own contracts, and here's the trap inside the trap: the natural IT instinct after any acquisition is to merge environments, and merging a non covered estate into ULA infrastructure quietly converts integration work into unlicensed use, inside an agreement everyone believes is unlimited. The wrong answers: A is the whole company assumption, and transaction teams repeat it right up until an audit corrects them. C confuses their paper with yours, retiring the acquired company's contracts changes nothing about whose definition covers the deployments. D overstates the problem, though note carefully: Oracle will happily propose exactly that, a new or expanded ULA, as the solution to a gap that one sentence at entry would have closed. The acquisition clause: cheapest line in the markup, most expensive line to lack. Now, the exit, written early.
Writing the exit at entry, four certification mechanics that belong in the signature draft. One, the counting method: what counts, installed and running, and from which data sources, with your own verified inventory named as acceptable evidence. That last phrase is session eight's audit clause lesson transplanted: the difference between defending your count from your records and defending it from whatever tooling Oracle proposes at exit. Two, cloud counting, last session's modern classic, closed with one plain sentence: deployments in public cloud environments count at certification. Every month of migration during the term makes that sentence worth more, and it is only purchasable today. Three, the timeline: a defined certification window, a defined period for Oracle to respond, and an agreed process if the parties disagree about the count. Undefined timelines favor whoever benefits from the clock, and at exit, that's not you. And four, the separation sentence: certification is a declaration process, not an audit trigger, and the paper should say so. Language separating the two keeps the fear gear, gear two of the renewal machine, out of your exit entirely, because the threat that review will be careful lands differently when the contract defines what review is. Four mechanics, maybe six sentences of drafting. At entry they're concessions Oracle can make cheaply. At exit they're the negotiation itself. Now, the money.
Pricing the unlimited period, and here's what makes it strange: the fee has no list price, no benchmark table, no public comparables. Session nine gave you discount bands for ordinary deals; nothing like that exists for ULAs. The fee is anchored entirely by preparation, yours or theirs, whoever brings the only number in the room. Five lenses. The forecast anchor: session eleven's method, honest growth priced à la carte at benchmark discounts. That ceiling is the entire negotiation, the fee lands under it or the ULA loses to buying as you grow. The annuity lens: fee times twenty two percent, forever, compare decades, never terms, a lower fee outranks nearly every other concession because it shrinks a perpetuity. The legacy fold in, watch this one closely: existing support streams often merge into the ULA's annuity, and fold ins are where the support base quietly ratchets, audit the before and after number line by line. The term choice: three years standard; a longer term dilutes the fee per year, which sounds attractive, but it also delays the exit skill your estate needs to learn and extends the annuity clock. And the walk away: à la carte purchasing at session nine discounts remains available every single day of the negotiation. The ULA must beat it, or the ULA is branding. Oracle prices these agreements on two inputs: your fear of counting, and your enthusiasm for the word unlimited. Arrive with the forecast, and both levers stop working. Choreography next.
Running the entry negotiation, five rules, and the first two come straight from session ten with the stakes raised. Rule one, paper before price: the list, the definitions, and the certification mechanics agreed before the fee conversation opens. In a ULA this matters double, because the certification language you win here is what the exit negotiation will be conducted with, three years from now, by people who may not be in the room today. Rule two, one markup, all blocks: the entire entry checklist as a single redline, definitions, cloud sentence, caps, session eight clauses, delivered together, valued internally. Rule three, spend the moment: a seven figure signature is maximum leverage, the top of session eight's list. The renewal cap and the audit terms that took escalation in session ten's ordinary deal are reachable here, and the certification mechanics, which exist nowhere on Oracle's standard concession menu, move only under this much weight. Rule four, model three exits before signing: certify low, certify high, renew, each with rough numbers. A deal you'd sign only under the happy exit is not a deal, it's a hope with a wire transfer attached. And rule five, keep the alternative alive: the à la carte path stays priced, current, and visibly on the table until the moment of signature. Unlimited is a convenience, not a necessity, and everyone in the room should be able to feel that you know it. Final check, on the number itself.
Knowledge check three. Your honest three year forecast, priced à la carte at benchmark discounts, totals three point two million dollars. Oracle proposes a four point two million dollar ULA fee. The right response? A, sign, unlimited is worth a premium over any forecast. B, sign if the term can be extended to five years. C, counter from the forecast: the fee must land under the à la carte ceiling, or the ULA simply loses to buying as you grow. Or D, refuse ULAs on principle and buy à la carte regardless of the numbers. Pause here. What's the ceiling, and who set it?
The answer is C, and notice who set the ceiling: you did, with homework, before the negotiation started. That's what makes it a ceiling instead of an opinion. At four point two against a three point two ceiling, the ULA pays a million dollar premium for flexibility that only pays off if the forecast is badly wrong in the happy direction, and then it pays twenty two percent of that premium annually, forever, roughly two hundred twenty thousand a year of pure annuity difference. The counter writes itself, and it's three sentences: here is our forecast, here is the à la carte cost, the fee lands meaningfully below that number or we buy as we grow. Nothing adversarial about it; it's just arithmetic, delivered calmly, by the only party in the room holding a number. The wrong answers: A pays seven figures for a word, and yes, upside insurance has some value, but a million plus its perpetuity is a lot of premium for insurance against good news. B fixes the wrong variable, stretching the term dilutes the fee per year while extending the annuity clock and postponing the exit skill, it makes the number look better and the deal worse. And D throws away the genuinely good version: session eleven showed the aggressive growth scenario landing at fifteen thousand a processor, less than half benchmark, when growth honestly outruns the ceiling, the ULA wins, and refusing it on principle costs real money. The forecast opens the conversation, the ceiling closes it, and whoever brings the only number anchors the room. Let's see the finished term sheet.
The term sheet, before and after, five rows. The product list: nine products in the draft, the full stack; four in the negotiated version, the growth products, precisely named, WebLogic having survived the honest forecast question, Analytics, GoldenGate, and the stowaway having not. The fee and its annuity: four point two million dropping to two point nine, and watch what travels with it, support falls from nine hundred twenty four thousand to six hundred thirty eight thousand a year, with the uplift capped at three percent. Over a decade, that row alone is worth more than four million dollars. Customer and M&A: signing entity only in the draft; all majority owned affiliates with automatic acquisition inclusion in the markup, the sentence from knowledge check two, purchased at the only moment it's cheap. Certification: the draft's standard clause, silent on cloud; the negotiated version counts cloud at exit, accepts your own inventory as evidence, defines the timeline, and separates certification from audit. Six sentences that will conduct the entire exit negotiation, three years early. And the session eight block: absent from the draft, obviously; present in the markup, forty five day audit terms, divestiture transition, price holds on the products that stayed off the list. The note under the table is the session's whole argument: the negotiated column is one prepared markup at one leverage moment. The draft column is what signing the proposal as sent would have locked in, for three years, with the window welded shut. That's the working session, ULA edition.
Session twelve in three sentences. One, a ULA has exactly one real negotiation, the weeks before signature, so the entry markup is the highest leverage document in the module, and everything on it follows the same rule: a sentence now or a fortune later. Two, the product list is scoped to real growth and nothing else, the definitions are sized for an unlimited grant, affiliates, acquisitions, territory, and the certification mechanics, cloud counting first among them, are written into the entry paper while they're cheap. Three, the fee is anchored by your own forecast and the à la carte ceiling, compared across decades because it carries a twenty two percent perpetuity on its back, and the arithmetic, not the word unlimited, gets the final say. Next session, we live inside the agreement: deployment tracking during the unlimited years, handling acquisitions and divestitures mid term, the cloud migration question in real time, and building the records that certification will demand, three years before it demands them. The entry paper is signed. Now we have to operate it. See you in session thirteen.
Homework, about an hour, and it works whether or not a ULA is anywhere on your horizon, because the exercise is the skill. One, draft your list: if a ULA were proposed tomorrow, which products belong on it, from real growth plans, and which would be padding? Write both columns; the padding column teaches you more. Two, build the ceiling: price that growth à la carte at session nine discounts, support included, and write the number down. That number anchors any fee conversation your company ever has, and it costs an hour to own. Three, check the definitions: against your actual org chart and your actual deal pipeline, what would the customer definition and the acquisition clause need to say to cover the next three years of corporate reality? Four, write the cloud sentence, one sentence, that would make your cloud deployments count at certification. Keep it somewhere safe. Depending on your migration plans, it may be the most valuable sentence you ever draft. And five, model three exits: certify low, certify high, renew, rough numbers each. If one of the three frightens you, that's the one that needs a plan, and better to meet it on paper today than in a conference room in year three. That's the hour. See you in session thirteen.
Five reads, all free on redress compliance dot com. First, the Oracle ULA negotiation playbook, the entry checklist in full reference form, block by block, with wording. Second, Oracle database ULA negotiation, entry terms for the database ULA specifically, the version most companies actually sign. Third, negotiating and managing an Oracle PULA, the perpetual variant's ten contract traps, worth reading now because PULA proposals often appear as counters during entry negotiations, before session fifteen prices the no exit option properly. Fourth, field tested Oracle negotiation strategies, the leverage moment doctrine applied at scale, today's rule five with war stories. And fifth, Oracle licensing in M&A due diligence, the acquisition clause from the transaction side, what corporate development wishes licensing had negotiated. That's session twelve. One negotiation window, five blocks, a ceiling you set yourself, and a term sheet where every improved row was bought with preparation. The unlimited period starts next session, and with it the part most companies get wrong: the living. See you in session thirteen.