Named User Plus looks cheaper for a small finance user base, but the 25-per-processor minimum and indirect access rules routinely wreck that assumption. This page shows exactly where the crossover sits and what to license before Oracle counts it for you.
Named User Plus looks cheaper for a small finance user base, but the 25-per-processor minimum and indirect access rules routinely wreck that assumption. This page shows exactly where the crossover sits and what to license before Oracle counts it for you.
Oracle sells Hyperion Planning Plus and Hyperion Financial Management Plus under two metrics: Application User (a Named User Plus variant, sometimes shown as NUP on older paper) and Processor. For a defined finance team, the user metric is almost always cheaper. The trap is that the user metric carries a per-processor floor, and most finance teams license to headcount and ignore the floor. That single mistake is the most common Hyperion audit finding we see, and it is entirely avoidable.
The rule you must apply on every server is simple: you license the larger of your actual authorized user count or the Named User Plus minimum computed from the hardware. Not the smaller. Not the one that suits your budget. The larger. If you take one thing from this page, take that. For the broader entitlement picture, read our Oracle Hyperion EPM on-premise licensing guide.
You license the larger of actual authorized users or the per-processor minimum. Costing to headcount alone is how estates end up under-licensed on paper before an auditor ever arrives.
These are the current US Public Sector price list figures (May 21, 2026). Commercial list prices are close, and both are heavily discountable in a negotiated deal. Use these as anchors, not as what you should pay. Annual support runs at roughly 22% of net license, the figure Oracle has held on Hyperion for years.
| Product | Metric | List per unit | Annual support | Part |
|---|---|---|---|---|
| Hyperion Planning Plus | Application User | $3,500 | $770 | L61277 |
| Hyperion Financial Management Plus | Application User | $5,200 | $1,144 | L61269 |
| Hyperion Tax Governance | Application User | $4,500 | $990 | L101673 |
| Data Relationship Steward | Application User | $5,800 | $1,276 | L72025 |
Note there is no Processor line in this table because the Processor SKUs are priced differently and, for Planning and HFM's typical finance-department footprint, they rarely win. That is the point of the comparison: the user metric is not just cheaper per unit, it maps to how these applications are actually used. Oracle's own guidance says Processor licensing is aimed at organizations with hundreds or thousands of users, and NUP suits smaller defined teams. In 25 years negotiating with this vendor, we have seen very few Hyperion Planning or HFM estates where the user population genuinely justifies Processor.
Here is where finance teams lose money and create exposure. The Named User Plus minimum for Hyperion is 25 licenses per processor, and "processor" is a licensing count derived from your hardware, not the raw socket or core number. The sequence is fixed and non-negotiable:
Worked example: an 8-core Intel server at a 0.5 core factor equals 4 processors. Four processors times 25 equals a 100-license NUP minimum. If only 30 analysts use Planning on that box, you still owe 100 Application User licenses, not 30. At $3,500 list that gap is $245,000 of license plus roughly $54,000 a year in support that many finance teams never budgeted for.
Two subtleties that catch people. First, the core factor never reduces your user count. It only sets the processor number the minimum is calculated from. If your actual user count exceeds the floor, the core factor is irrelevant to you. Second, if you push Hyperion into an Authorized Cloud Environment, the Core Factor Table is expressly excluded and you count vCPU instead. Do not carry a 0.5 assumption into a cloud business case. See our Hyperion to EPM Cloud migration cost analysis for how the metric changes entirely in the cloud.
The generic Oracle Database crossover is well documented: at roughly 50 users per processor-equivalent, Processor starts to win. Below that, the user metric is cheaper. But Hyperion's Plus SKUs are priced differently from the database, so do not blindly import the 50-user rule. Our detailed treatment of the underlying math lives in the Named User Plus or Processor metric decision reference. For Hyperion specifically, the practical test is:
A concrete comparison for Planning Plus at list. Say a single 8-core Intel server (4 processors after core factor):
| Scenario | Actual users | NUP floor | Licensed quantity | Approx. list |
|---|---|---|---|---|
| Small team | 20 | 100 | 100 (floor wins) | $350,000 |
| Mid team | 80 | 100 | 100 (floor wins) | $350,000 |
| Larger team | 150 | 100 | 150 (users win) | $525,000 |
| Growing / indeterminate | 400+ | 100 | evaluate Processor | model both |
The table makes the real lesson obvious: below the floor, adding users up to 100 costs you nothing extra because you already owe 100. Many estates that think they are 20-user shops are actually paying for 100 whether they use them or not. That is a licensing entitlement worth reconciling against actual access, and a fact worth remembering when Oracle tries to sell you more.
Below the per-processor floor, extra users are free because you already own them. Reconcile the floor against actual access before you buy a single additional license.
The metric choice is only half the exposure. The other half is who and what counts as a user. Oracle's counting rules are broader than most finance teams assume, and these are the ones we see cited in audit findings repeatedly.
Named User Plus is not concurrency, and it is not "people who log in directly." It counts every user or device authorized to access the software, including through an intermediate application. If a web front end or a reporting tool retrieves Essbase or Planning data for end users, Oracle's position is that each of those end users needs a license even though they never see a Hyperion login screen. Multiplexing does not launder the requirement. This is the single most expensive misread in the whole area.
Hyperion Planning includes an embedded Essbase engine, but only for use with Planning's own data. Build a separate Essbase cube or an application unrelated to Planning and you are using Essbase outside your entitlement, which requires a separate Essbase license. HFM behaves differently: it does not use Essbase as its primary data store (it runs on a relational database), so the Essbase counting rules do not apply to HFM data in the same way. Do not assume the two products carry the same options exposure. Our Essbase standalone versus bundled licensing analysis unpacks this trap in full.
Bundled suites such as Hyperion Planning Plus and Hyperion Financial Close Suite package several modules under one SKU, but the license quantities for each included component must match and usage is restricted to the suite's scope. Each product still needs its own license: buying one module never grants rights to another. Finance teams that treat a suite as an all-you-can-eat license routinely deploy modules they never bought.
Hyperion ships with an Oracle database and WebLogic underneath it. The restricted-use rights that come with Hyperion cover only the Hyperion workload, and any additional use of that database or WebLogic requires its own license. This is a classic hidden liability that surfaces mid-audit. We cover it separately in the database and WebLogic licensing buried in your Hyperion stack.
Hyperion license agreements are inflexible once signed. You pay for what you license, and reducing quantities later is difficult to impossible. That asymmetry means your leverage is entirely at the point of purchase and renewal. Use it. Here is the sequence we run for clients:
If your estate is already on maintenance and you are weighing whether to keep paying Oracle support at all, the metric analysis feeds directly into that decision. See our Hyperion sustaining support end-of-life decision and the third-party support decision for the downstream economics. And if you are already in an audit, the counting rules above are exactly what you will need to defend, covered in defending a Hyperion audit.
For Hyperion Planning and HFM in a finance-department footprint, the Application User (NUP) metric is almost always the right and cheaper choice. But the metric is only cheap if you compute the per-processor minimum correctly and count indirect users honestly. The failure mode is not choosing the wrong metric. It is choosing the right metric and then under-counting the floor, which leaves you under-licensed on paper and exposed the moment Oracle audits. Compute the floor, license the larger of floor or users, keep modules inside scope, and treat every hardware change as a re-modeling event. That discipline turns Hyperion licensing from a recurring audit liability into a predictable, defensible line item.
The minimum is 25 Named User Plus (Application User) licenses per processor of the server running the component. You calculate the processor count from your cores using Oracle's Core Factor Table (0.5 for Intel and AMD x86), round up to the next whole number, then multiply by 25. You license the larger of that floor or your actual authorized user count.
Processor typically wins only when your user population is large, growing, or genuinely indeterminate, for example a consolidation platform read by hundreds of business users. For a defined finance team using Planning or HFM, the Application User metric is almost always cheaper. Because Hyperion Plus SKUs are priced differently from the database, model both totals directly rather than assuming the generic 50-user database crossover applies.
No. The core factor only sets the processor number that the 25-per-processor minimum is calculated from. It never reduces your actual user count. If your authorized users already exceed the minimum, the core factor has no bearing on what you owe.
Yes. Named User Plus counts every user or device authorized to access the software, including access through an intermediate web or reporting application. If a front end pulls Essbase or Planning data for end users, each of those end users needs a license even if they never log into Hyperion directly. Multiplexing does not remove the requirement.
Rarely and with difficulty. Oracle Hyperion license agreements are inflexible once signed, so reductions after purchase are hard to obtain. Your leverage sits at purchase and renewal, which is why you should compute the per-processor floor and reconcile actual access before committing to any quantity.
Only if you use it outside Planning's scope. The embedded Essbase engine is licensed for use with Planning's own data. If you build separate Essbase cubes or applications unrelated to Planning, that is out of entitlement and requires a standalone Essbase license. HFM does not use Essbase as its primary data store, so this rule affects Planning more than HFM.
Oracle sells the same database per user and per processor. The 25 NUP per processor minimum, the 50 user break-even, and how to choose the metric that fits the workload.
Gated with a work email on the download page. No sales follow up you did not ask for.
Get the White Paper →500+ enterprise clients. 11 vendor practices. Industry recognized. One conversation can change what you pay for the next three years.
One buyer side briefing a week. Renewal signals, audit moves, and the levers that work. No vendor spin.