Subscriptions renew. The number that matters is the price in year four.
An Oracle Fusion Cloud ERP negotiation is a single event that sets a cost base for six years, and four things decide it: the metric, the ramp, the renewal uplift, and a short list of contract clauses. All of them are open until you sign and effectively closed afterwards. The opening proposal is a position rather than a price, and the value leaks in the term nobody negotiates.
Prepared by Redress Compliance · August 10, 2026 · Oracle advisory. Based on 25 to 35 Oracle Fusion Cloud negotiations, 2024 to 2025.
Executive summary
Median improvement from first quote to signature ran 18 to 34 percent once the metric and the uplift were both worked. The opening proposal is a position, and it almost never held in our file.
But the two levers that produced that movement are not discount levers: they are the counted population and the price trajectory, and a team that negotiates hard on percentage while leaving both untouched has won the smallest argument available in the deal.
Deals signed without a renewal uplift cap saw 8 to 12 percent annual increases that erased the first term discount within three years. A cap is a non standard term requiring deal desk approval, which means it has to be tabled in the first two weeks rather than the final week.
Teams that raised it in the last three weeks were told approvals could not be obtained in time, and that was often genuinely true rather than a tactic. Timing is the whole difference between a cap you get and a cap you are refused.
The metric was mismatched to usage in over half the estates. Hosted Named User counts every individual authorised to use the service whether or not they log in, which suits deep use by a contained finance and procurement team.
Employee metrics count a population regardless of direct use, which suits broad, light touch deployment such as expense entry across the workforce.
Buyers bought the wrong one in both directions, and the trap is letting Oracle pick the metric that maximises the count rather than the one that reflects usage. Settle the metric before anyone discusses discount.
The ramp is a contractual quantity by year, not a consumption forecast. You pay the ramped quantity whether or not the users exist, and the full contracted quantity rather than the consumed quantity sets the renewal baseline, which is how ramps get used to inflate the renewal price.
The same applies to modules handed over at no charge in term one: they carry a quantity into the renewal quote and get priced at list when the free period ends. Quantities also ratchet up mid term as soon as you exceed them and only come down at renewal, and only if you negotiated a reduction right.
Choosing the metric before anyone talks discount
| Metric | What it counts | Where it wins | The trap |
|---|---|---|---|
| Hosted Named User | Each individual authorised, logged in or not | Deep use by a contained finance and procurement team | Counting occasional users at full rate |
| Employee based | A population regardless of direct use | Light, broad use such as expense entry across the workforce | Counting a whole headcount for a narrow deployment |
| Either, badly defined | Whoever the definition captures | Nowhere | Two orders saying Employee can count different people |
The definition beats the label, and the definition belongs in the ordering document rather than in a conversation.
Two orders can both say Employee and count entirely different populations, so specify which legal entities are in scope, whether contractors and seasonal staff count, and the measurement date that determines the number.
One structural fact applies to every Fusion metric regardless of which you pick: quantities ratchet up during the term as soon as you exceed them and they only come down at renewal, and only if you negotiated a reduction right in advance.
That asymmetry is why the metric decision is worth more than the discount conversation that follows it. The module rate card sits in the ERP Cloud pricing brief and the applications landscape in the Fusion applications guide.
Where the value leaks
- The renewal uplift. Uncapped, it runs 8 to 12 percent annually and erases the first term discount within three years. Table the cap in the first two weeks, because deal desk approval cannot be obtained in the last three.
- The ramp schedule. A contractual quantity by year rather than a forecast, so you pay it whether or not the users exist, and the contracted quantity sets the renewal baseline.
- Free modules in term one. They carry a quantity into the renewal quote and get priced at list when the free period ends, which makes them a deferred purchase rather than a concession.
- The mid term ratchet. Quantities rise as soon as you exceed them and fall only at renewal, and only with a negotiated reduction right, so the entry number matters more than it appears.
- Legacy support overlap. Cancelling E Business Suite or PeopleSoft support at go live can reprice the support lines you keep, so sequence the termination inside the same negotiation rather than after it.
The Oracle CIO complete playbook
The five year plan to control Oracle spend, including the subscription renewal arithmetic, the metric decision, and the clauses that have to be tabled early.
Get the white paper →Running the negotiation in the right order
Settle the metric first, because it sets the counted population and the counted population sets the bill for the whole term, and no discount recovers a population that was never the right one.
Table the clause asks second, in the first fortnight rather than the final one: a renewal uplift cap is a non standard term that needs deal desk approval, and teams that raised it in the last three weeks were told the approvals could not be obtained in time, which was frequently accurate.
A reduction right at renewal belongs in the same early conversation for exactly the same reason.
Work the ramp third, treating it as a contractual quantity schedule rather than a consumption plan: you pay what the schedule says, the contracted quantity rather than the consumed quantity sets the renewal baseline, and a ramp built on optimism therefore inflates the renewal price twice over.
Discount comes last and it is the smallest of the four, which is the opposite of how most negotiations are staffed. Two Oracle counters are worth anticipating.
Free modules in term one look like a concession and behave like a deferred purchase, because the quantity carries into the renewal quote and prices at list when the free period ends.
And Support Rewards, which offsets on premises technical support against OCI infrastructure and platform spend, does not apply to Fusion SaaS spend at all, so it should never be allowed into the Fusion business case.
Finally, sequence any legacy E Business Suite or PeopleSoft support termination inside the same negotiation, since cancelling at go live can reprice the support lines you intend to keep. The renewal side of Oracle subscriptions sits in the renewal strategy guide.
- Percentile standing for your exact deal size and industry, from real closed transactions
- Scenario simulation before the call: test alternative terms and see the financial impact of each
- A negotiation playbook, talking points, and a two page executive brief on day one
What we saw across Fusion ERP negotiations, 2024 to 2025
Across roughly 25 to 35 Oracle Fusion Cloud negotiations we advised on between 2024 and 2025, the first quote almost never held, and four patterns recurred:
Median improvement achieved once the metric and the renewal uplift were both worked, rather than the discount alone.
The increase on deals signed without a cap, which erased the headline first term discount inside three years.
The renewal uplift was where value leaked, with uncapped deals seeing 8 to 12 percent annual increases. The user metric was mismatched to usage in over half the estates, in both directions.
Ramp schedules were used to inflate the renewal baseline, because the full contracted quantity rather than the consumed quantity set the renewal price. And the clause asks arrived too late, with teams raising the uplift cap in the last three weeks told that approvals could not be obtained in time.
The buyer side move is to fix the metric, table the clauses in the first fortnight, treat the ramp as a quantity commitment, and argue discount last. Fusion is a subscription, and subscriptions renew.
Your first five moves
- Settle the metric before anyone discusses discount, and write the definition into the ordering document: entities in scope, contractor and seasonal treatment, and the measurement date.
- Table the renewal uplift cap in the first two weeks, because it is a non standard term needing deal desk approval that genuinely cannot be obtained in the final three.
- Negotiate a reduction right at renewal in the same early conversation, since quantities ratchet up mid term as soon as you exceed them and otherwise never come back down.
- Treat the ramp as a contractual quantity schedule, not a forecast, because you pay it regardless and the contracted quantity sets the renewal baseline.
- Sequence any legacy support termination inside the same negotiation, as cancelling at go live can reprice the support lines you keep. The Oracle practice runs the metric analysis and the negotiation with you.
Frequently asked questions
Which Fusion ERP user metric should we choose?
The one that matches how broadly each module is actually used, settled before anyone talks discount. Hosted Named User counts each individual authorised whether or not they log in, suiting deep use by a contained team.
Employee metrics count a population regardless of direct use, suiting light, broad deployment such as expense entry across the workforce.
Why does the metric definition matter more than the label?
Because two orders can both say Employee and count entirely different populations. Get the definition written into the ordering document: which legal entities are in scope, whether contractors and seasonal staff count, and the measurement date.
Without that, the label tells you nothing about the number you will actually be billed on.
When should the renewal uplift cap be raised?
In the first two weeks. A cap is a non standard term that requires deal desk approval, so it needs time in the process. Teams that raised it in the last three weeks were told approvals could not be obtained in time, and that was often genuinely true rather than a negotiating tactic.
Late is functionally the same as never here.
What happens without an uplift cap?
Deals signed without one saw 8 to 12 percent annual increases in our file, which erased the headline first term discount within three years. Fusion is a subscription and subscriptions renew, so the number that matters is not the first term discount.
It is the price in year four, and only the cap governs that.
Is a ramp schedule a forecast?
No. It is a contractual quantity by year, and you pay the ramped quantity whether or not the users exist.
Worse, the full contracted quantity rather than the consumed quantity sets the renewal baseline, so an optimistic ramp inflates the price twice: once during the term and again at the renewal that prices from it.
Are free modules in term one a concession?
They behave like a deferred purchase. Modules handed over at no charge carry a quantity into the renewal quote and get priced at list when the free period ends, so the concession is really a commitment to buy at an unnegotiated rate later.
Price them at renewal rates when evaluating the offer rather than at zero.
Do Oracle Support Rewards apply to Fusion?
No. Support Rewards offsets on premises technical support against OCI infrastructure and platform spend, and it does not apply to Fusion SaaS spend.
It should not be allowed into the Fusion business case, because including it inflates the apparent value of a deal by a benefit the deal does not carry.
What about legacy E Business Suite or PeopleSoft support?
Sequence the termination inside the same negotiation. Cancelling legacy support at go live can reprice the support lines you intend to keep, through the same matching service level and repricing mechanics that govern any Oracle support reduction.
Handled afterwards it becomes a separate and weaker conversation with no leverage attached.
