The loop that pays your support bill with your cloud bill, worked to the last cent. 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 nine of thirty, and the session three earlier threads have been building toward. Session one gave you the headline, twenty five cents per OCI dollar against the support bill. Session seven made you size consumption honestly, and session eight left you holding a support annuity that BYOL obliges you to keep paying. Today the threads pay each other off: Oracle Support Rewards, the loop that connects your cloud bill to your support bill, worked to the last cent. By the end you'll be able to compute exactly what your estate earns, know the expiry clock nobody sets an alarm for, and, maybe most valuably, recognize the negotiation move where a rep tries to sell you this program as a concession. It's the closest thing to free money in the Oracle relationship, and like all free money, the terms matter. Let's read them.
Five takeaways. One, the loop, precisely: how a dollar of consumption becomes twenty five or thirty three cents off the tech support bill, step by step, because precision here is what stops both the over claiming and the under claiming I see in equal measure. Two, the rules: what earns, what redeems, what expires, the fine print that decides whether the arithmetic actually lands in your favor. Three, the worked case: a real shaped estate taken from a two million dollar support bill to zero, every number on the table. Four, the operations: accrual tracking, redemption cadence, the expiry alarm, and the internal budget question that kills this program more often than any rule does. And five, the strategy: why Oracle pays you to consume, what that means for lock in, and how to spend the loop at a renewal without letting it be spent against you.
The loop, three steps, and get the direction right because half the confusion about this program is directional. Step one, consume: Universal Credits consumption on OCI services earns rewards. Twenty five cents per dollar consumed, thirty three cents if you hold an active unlimited license agreement. Mark the verb: consumption earns. Not commitment, not the contract signature, the actual metered burn. Session seven's forfeit lesson thus gets a sequel: an oversized commitment forfeits twice, once as credits, once as the rewards those credits never earned. Step two, accrue: rewards accumulate monthly as consumption happens, visible in the console right next to the billing data, denominated in dollars, belonging to the account. And step three, redeem: accrued rewards apply against Oracle tech support invoices, the twenty two percent annuity on your on premises licenses, potentially all the way to zero. Against tech support. Not against SaaS subscriptions, not against the cloud bill itself, not against new purchases. It's a rebate aimed at one specific target, and here's why the aim matters to you: the bill session eight's BYOL obliged you to keep paying is the exact bill this program pays down. BYOL and Rewards are one system, and estates that model them together, third column and rebate, price their cloud correctly. Estates that don't leave one of the two on the table.
The eligibility rules, five of them, each one a place real money gets found or lost. Rule one, Universal Credits earn: the program keys to UC consumption, and Pay As You Go estates don't accrue. If the Rewards arithmetic is material for your estate, that's one more argument for graduating from PAYG to a commitment, properly sized, and notice how the course's threads keep braiding: the commitment discipline and the rewards math argue for each other. Rule two, tech support redeems: on premises database and middleware annuity, that's the target, and SaaS subscriptions and cloud fees sit outside the loop entirely, we'll drill that in the first check. Rule three, and set an alarm before this sentence ends: rewards expire. Twelve months from earning, then they lapse. An unredeemed reward is a forfeit with better branding, and estates with a redemption backlog are watching free money time out monthly. Rule four, third party marketplace spend rides your commitment but doesn't earn the same way, so the console's accrual line is the truth, reconcile it against expectations rather than assuming every committed dollar earns. And rule five, the ULA bonus: thirty three cents instead of twenty five while an unlimited agreement is active. That's Oracle paying ULA customers a premium to build cloud habits before certification, and the strategic read of that comes later, but the tactical read is simple: if you hold a ULA, take the thirty three cents, and model the drop back to twenty five into your exit economics.
First check. An estate consumes two hundred thousand a month of Universal Credits, pays one point four million a year in tech support, and eight hundred thousand in Fusion SaaS subscriptions. At the standard rate, what do Rewards do to those two bills? A, six hundred K a year accrues and can be split across both bills. B, six hundred K accrues and applies against the one point four million tech support bill only, taking it toward eight hundred K, while the SaaS bill is untouched. C, nothing accrues, two hundred K a month is below the program floor. Or D, six hundred K accrues and reduces the OCI bill itself. Pause here. Two hundred K times twelve times a quarter. And what does a reward redeem against?
The answer is B. Two hundred thousand a month is two point four million a year of consumption, times twenty five cents is six hundred thousand of accrual, and the redemption target is the tech support annuity, alone. The one point four million drops toward eight hundred thousand, and the Fusion subscription doesn't move, because SaaS sits outside the loop, full stop. There's no meaningful program floor, small consumption accrues small rewards just fine, so C is noise. And D is the misreading I most want to retire today, because it changes how you think about the program: Rewards never touch the cloud bill. They're not a cloud discount. They're a rebate, aimed at the annuity, funded by consumption. The effective outcome for this estate: total Oracle spend drops six hundred thousand a year, without a single negotiation, for the price of redeeming on time. Which, as we'll see in the operations slide, is a real price that real estates fail to pay.
Now the worked case, the one to keep. The estate: two million a year of tech support, the twenty two percent annuity on a substantial on premises license base. The cloud side: eight million a year of measured Universal Credits consumption, sized session seven style, actually burning, not just committed. The accrual: eight million times twenty five cents is two million a year of rewards. Redemption: applied to every tech support invoice as it lands, and the two million dollar support bill goes to zero. Not reduced. Zero. And notice the restatement in the last line of the table, because it's how the CFO will want to hear it: with support fully offset, every OCI dollar effectively cost seventy five cents, a built in twenty five percent that stacks under whatever discount you negotiated on the credits themselves. Now the identity that generalizes the whole table, and it's one line: consumption times rate, held against the bill. At twenty five cents you zero the bill when consumption is four times the bill. At thirty three cents, three times. Below breakeven the program still pays, just partially, like the check we did a minute ago. Four x at standard rate, three x with a ULA. That identity prices every scenario in this session, and you can do it on a napkin in front of anyone. Speaking of which, let's watch our advisor do exactly that.
Guest analyst The most effective five minutes I spend with a CFO is the Support Rewards napkin. A client was evaluating their Oracle exit: two million of support they resented, a third party support quote at half price, and a real appetite to leave. The napkin had three lines. Line one: your migration plan already puts seven million of workloads on OCI over two years, priced against AWS. Line two: seven million times twenty five cents is one point seven five million a year against the two million bill. Line three: so the OCI path does not just compete with AWS on rate, it quietly pays off almost your entire support bill, which the AWS comparison your team built treats as fixed on both sides. That was the meeting. The AWS TCO was rebuilt with the offset in, OCI won by a distance it had previously lost by, and the third party support conversation ended, because a bill offset to two hundred fifty thousand is not worth the patch limitations. Two warnings from that engagement, though. First, the offset is contingent: the year consumption dips, the bill comes back, so we modeled the dip scenario before signing anything. And second, we still negotiated the credit discount as if Rewards did not exist, because they are a program, not a concession, and the rep absolutely tried to count them as one. Separate ledgers. Every time.
Three lines on a napkin that rebuilt a TCO and ended a third party support conversation. And catch both warnings, because they're the rest of this session: the offset is contingent on consumption, model the dip. And the rep tried to count the program as a concession, separate ledgers. We'll test both before we're done.
Check two, the napkin in your hands. A CFO asks: we pay one point two million in support, we're planning three million a year of OCI. Do we zero the bill? And should the answer change the migration business case? A, yes, zeroed, any OCI spend eliminates support. B, no: three million times twenty five cents is seven hundred fifty K, the bill drops to four hundred fifty K, and yes, the case changes, because seven hundred fifty K of annual offset belongs in the migration's business case as hard savings. C, no, and irrelevant, Rewards are a rounding error. Or D, yes, if the three million is committed rather than consumed. Pause. Consumption times rate against the bill. And who should be told?
The answer is B, both halves. The identity: three million times twenty five cents is seven hundred fifty thousand, against a one point two million bill, leaving four hundred fifty. Zeroing would need four point eight million of consumption, four x the bill, which this plan doesn't reach. And the second half is the half that changes careers: seven hundred fifty thousand a year of hard, contractual, published program offset belongs in the migration business case, and any OCI versus AWS comparison that omits it is wrong by exactly that amount, in Oracle's disfavor, which is the advisor's story from thirty seconds ago playing out in reverse. C dismisses three quarters of a million dollars a year as rounding, which speaks for itself. And D, look at it carefully, because it's session seven wearing a new costume: committed doesn't earn, consumed earns. A three million commitment consumed at two million accrues five hundred K, not seven fifty, and forfeits a million in credits besides. The napkin only works on measured numbers. Everything in this course only works on measured numbers. That's arguably the course.
The operations, five habits, because a rebate you don't operate is a rebate you don't receive. One, track the accrual: the console shows rewards earned next to consumption, and it joins the session six monthly review as a standing line, earned, redeemed, expiring. Three numbers, one glance. Two, redeem on cadence: apply rewards to every tech support invoice as it lands. The twelve month expiry means a redemption backlog isn't a savings account, it's a countdown, and I've reviewed estates sitting proudly on six figure balances that were sixty days from lapsing. Three, alarm the expiry: a standing alert ninety days before any accrued balance times out. The program's only genuine failure mode is free money quietly expiring, and it's entirely preventable with one calendar entry. Four, the one that actually kills the program: route the credit. The rebate lands on the support renewal, which is usually an infrastructure or procurement budget, but the consumption that earned it came out of cloud budgets. If nobody settles whose budget benefits, the cloud team has no incentive to care about accrual, the support owner doesn't know it exists, and the program starves in a turf war. One paragraph, agreed with finance, once. And five, reforecast when support changes: terminations, ULA exits, and BYOL moves all change the redemption target, and a loop is only worth the bill it offsets. Both sides of the identity, recomputed together, every time either moves.
Now zoom out, because understanding why this program exists tells you exactly how hard you can lean on it. Read one: the support base is the prize. Oracle's on premises support stream is among the most profitable revenue in enterprise software, and it's under siege, third party support firms halve it, cloud exits erase it. Support Rewards is the defensive masterstroke: it converts would be defectors into OCI consumers who keep paying support, technically, while Oracle rebates it back at a rate Oracle controls. Everyone stays in the tent. Read two: the lock in arithmetic, the advisor's first warning. An estate whose two million dollar support bill is fully offset has acquired a new switching cost, leave OCI, and the two million comes back. That's not a reason to refuse the program, it's a reason to model the exit scenario with the offset removed, before you celebrate the zero, so the day someone proposes leaving, the real price is already on paper. And read three, the negotiation footnote that pays for this session: Rewards are a published program, available to every Universal Credits customer, which means they stack on top of negotiated discounts and are never, ever part of one. Session five's benchmarks assume them. The rep who folds them into a total value package is anchoring you, and the next check is that exact conversation.
Last check, the anchoring move. In a commitment negotiation, the rep proposes: thirty percent credit discount, and we'll include Support Rewards, presenting it as a fifty five percent total value package. The correct response: A, accept, fifty five percent total value beats session five's benchmark. B, separate the ledgers: Rewards are a published program every UC customer gets, so the negotiated discount on the table is thirty percent, below benchmark for this deal size, and the discount conversation continues from there. C, reject Rewards to keep the negotiation clean. Or D, ask for fifty five percent discount plus Rewards on top. Pause here. Which part of that package did the rep actually concede?
The answer is B. Audit the package: the rep conceded thirty points of discount. The other twenty five points belong to a published program your estate would earn anyway, the same as every Universal Credits customer with a support bill on the books, including the ones who never negotiated at all. Folding table stakes into the concession count is textbook anchoring, and the counter isn't outrage, it's bookkeeping: two ledgers, program value on one side, negotiated value on the other, and then the calm sentence, so the offer on the table is thirty percent, and for this deal size the benchmark says more. The conversation continues from the real number. C, refusing the program to keep things clean, punishes yourself to make a point nobody receives, the money is real, take it. And D at least has the ledgers straight but mistakes the benchmark for a floor, fifty five plus Rewards isn't a position, it's a wish. The discipline is the advisor's closing line from the clip, and it's worth being the last words before the recap: separate ledgers. Every time.
Before we close, the four places this arithmetic changes some other negotiation, because the loop never acts alone. One, the support renewal itself: a bill fully offset by Rewards is still worth attacking, caps, right sizing, restructuring, all of the Mastery course playbook, because the offset is contingent on consumption and the bill underneath is forever. Negotiate as if the offset didn't exist; bank the difference when it does. Two, the platform decision: in any OCI versus AWS or Azure comparison for Oracle workloads, the offset is OCI exclusive money, and leaving it out biases the TCO by up to a quarter of your consumption. Module three's grand comparison carries it, and now you know why. Three, the third party support fork: Rimini style support halves the bill, but it exits the loop, no Oracle support means nothing to redeem against. At high OCI consumption, the Rewards offset can beat the third party discount outright, the advisor's client landed exactly there, so run both curves before choosing a fork this permanent. And four, the ULA exit: thirty three cents lives only while the ULA does. Time heavy consumption into the ULA window where each dollar earns more, and model the drop to twenty five into the exit economics, one more line in the certification spreadsheet the Mastery course taught you to build.
Session nine, three sentences. One: the loop is precise, Universal Credits consumption earns twenty five cents per dollar, thirty three with an active ULA, redeemable against on premises tech support only, expiring twelve months after earning, so redemption cadence is not optional. Two: the identity runs every model, consumption times rate against the bill, four x zeroes it at standard rate, three x with a ULA, and the offset is hard money that belongs in every migration case and every platform comparison your estate produces. Three: Rewards are table stakes that stack on negotiated discounts, never a concession, and the strategic price of a zeroed bill is a new switching cost you model before you celebrate. Next session closes module two: OCI cost governance, budgets, compartments, tagging, and the monthly operating rhythm that keeps the meter, the commitment, and this loop honest, the operational floor under everything the module built. After that, module three, and Oracle finally meets the other clouds. See you there.
Homework, about an hour, and it's an audit of free money. One, find the accrual: your Rewards balance in the console, earned, redeemed, and anything approaching expiry. If rewards have already lapsed, compute what they were worth, not to assign blame, but because that number funds the alarm you're about to set. Two, compute the identity: trailing twelve month consumption times your rate, against the tech support bill, and write down the coverage percentage. That's your estate's number, and you'll use it in three future sessions. Three, check the redemption cadence: who applies rewards to support invoices, and how long after accrual? If the answer is nobody knows, congratulations, you found this week's fix and it's worth real money. Four, correct one business case: find the most recent OCI versus anywhere comparison your estate produced and check whether the offset was in it. If not, reissue the math, politely, with today's identity attached. And five, settle the budget question: one paragraph, agreed with finance, on whose budget the redeemed rewards benefit. It's the least technical item on this list and the one most likely to determine whether the program works at all. Turf kills more rebates than arithmetic ever has.
Five reads before next session, all free on redress compliance dot com. First, the Oracle Support Rewards guide, today's session in reference form, the rules, the rates, and the redemption mechanics to check your console against. Second, the OCI licensing and cost guide, the consumption machinery the accrual keys off. Third, the Oracle BYOL comprehensive guide, because the annuity BYOL keeps alive is the loop's redemption target, and the two programs should live in one spreadsheet. Fourth, Oracle cloud negotiations, where the two ledger discipline fits in the wider negotiation playbook. And fifth, OCI versus AWS for Oracle workloads, the platform comparison the offset belongs in, and the perfect on ramp for module three, which is exactly where this course goes after one more session of governance. That's session nine. The loop is real, the identity fits on a napkin, and the ledgers stay separate. See you at session ten.