A decision priced at 150 percent to reverse is not a price decision

Third party support is usually evaluated the way a commodity is evaluated: line up the providers, compare the discount, pick the cheapest credible one. The reinstatement term makes that the wrong shape of analysis, because the mistake you can make here is not an expensive quarter, it is a five year estate you cannot easily undo.

Prepared by Redress Compliance · August 11, 2026 · Oracle advisory. Based on 20 to 30 third party support selections run or reviewed, 2024 and 2025.

• Start from your module and version list, not a provider shortlist, because coverage of that exact set decided 60 to 70 percent of the selections we scored.
• Read the indemnity language side by side, since it varied more between providers than price did and it governs the scenario the arrangement exists to handle.
• Ask for reference customers on your platform specifically, as those predicted satisfaction better than provider revenue did in every selection we reviewed.
• Test your own roadmap first, because the estates that struggled after the move were the ones still changing, and that is a fact about you rather than about the provider.
• Check what else terminates with the contract, including Support Rewards, since OCI credits that offset on premises support disappear along with the support they offset.

The reason a fifty percent saving does not make this a simple decision is the shape of the downside rather than its size. Most software decisions are recoverable at roughly the cost of having made them: you overpay for a term, you renegotiate at renewal, you move on. This one is not.

Oracle prices reinstatement at 150 percent of the last annual fee you paid, and no amount of money buys back the patches you did not receive while you were away, because there are no new Critical Patch Updates, no new tax and regulatory updates.

And no upgrade rights for the period after termination.

So a buyer who picks the wrong provider does not simply have a bad supplier, they have a bad supplier and an expensive, partial route back. That asymmetry is what should set the depth of diligence, and it explains why our selection data looks the way it does.

Price was almost never the deciding factor, because price is the recoverable variable and fit is not: coverage of the exact module and version set decided 60 to 70 percent of selections, reference customers on the same platform predicted satisfaction better than provider revenue.

And indemnity language moved the risk more than the price did.

There is a second constraint worth planning around before the provider conversation starts, because it changes the size of the decision rather than its quality.

You usually cannot move part of a license set, so the remainder has to be terminated or repriced, which means the scope of what leaves Oracle support is frequently larger than the piece you wanted to save money on.

Combine that with matching service levels and the practical result is that this is an estate level decision presented as a line item one. The single best predictor of how it goes is not in any provider comparison: the estates that struggled were the ones still changing, and stable estates were fine.

Run that test on your own roadmap first, because if the estate is mid programme the right answer may be to wait rather than to choose better. The provider market and the legal position are set out in the third party support guide.

Across roughly 20 to 30 third party support selections run or reviewed between 2024 and 2025, price was almost never the deciding factor and fit was:

Four patterns recurred: coverage of the exact module and version set deciding 60 to 70 percent of selections, indemnity language varying enough between providers to move the risk more than price did, reference customers on the same platform predicting satisfaction better than provider revenue.

And the estates that struggled after the move being the ones still changing.

The wider library sits in the Oracle practice.

Who supplies Oracle third party support?

Two providers operate at scale across database, middleware, and applications, with a long tail of specialists covering narrower product lines, regions, or regulatory jurisdictions. Provider scope changes, so read current disclosures rather than a comparison written two years ago.

What actually decides the selection?

Coverage of the exact module and version set, which decided 60 to 70 percent of the selections we scored. Price was almost never the deciding factor, because price is the recoverable variable in this decision and fit is not.

How much does leaving Oracle support save?

Providers typically position against your current Oracle bill, which runs at 22 percent of net license fees a year under Oracle's own policy, with the anchor set around a 50 percent reduction. The saving is real and large, and so is the constraint that comes with it.

What do you give up by leaving?

After termination there are no new Oracle Critical Patch Updates, no new tax and regulatory updates, and no upgrade rights. Those cannot be bought back later at any price, which is what makes the decision asymmetric rather than simply expensive.

What does it cost to go back?

Oracle's published technical support policies price reinstatement at 150 percent of the last annual fee you paid, and that only restores future support. The patches and regulatory updates you did not receive while away are not recoverable, so the route back is both expensive and partial.

Is a larger provider a safer choice?

A listed provider is easier to diligence, because audited financials, customer concentration, and litigation status are published rather than requested. That is a real advantage on a long contract, but it is not a quality signal for your product family, where a deep specialist often serves better.

Who should not move to third party support?

Estates that are still changing. The clearest predictor in our reviews is that estates mid programme struggled after the move while stable estates were fine. If your roadmap is active, the right answer may be to wait rather than to choose a different provider.