Phantom seats inflated the buyout quote 20 to 35 percent before the baseline was clean

A perpetual unlimited agreement has no certification event and no expiry, so every exit is a negotiated contract rather than a scheduled one. Across the exits we benchmarked, the settlement moved far more on the quality of the deployment baseline than on the headline support figure.

Prepared by Redress Compliance · August 15, 2026 · Oracle advisory. Based on 30 to 40 benchmarked PULA exits, 2024 to 2025.

• Baseline every entity, geography and cloud platform before the first exit conversation, mapped to the published counting rules rather than to your asset register.
• Strip the phantom seats, the decommissioned instances, the test estates and the duplicated records, because each one is a permanent support line if it survives into the contract.
• Frame on a corporate event where one exists, since a divestiture or carve out supplies a deadline Oracle cannot move and you cannot postpone.
• Keep at least two paths live, because a buyer with one option is a buyer with a price, as the exit playbook sets out in full.
• Treat the exit agreement as a fresh contract, setting the entitlement, the support floor, the assignment language and the audit clauses together.
• Write the settlement target before the first call, one figure for each path, approved by the CFO rather than discovered in the meeting.

The standard advice from the account team is that a buyout is the clean way out: take the count, convert it to a perpetual entitlement, and the unlimited right ends tidily. The mechanism is right and the sequence is wrong. In roughly six of every ten exits we rebuilt, the count carried phantom seats that inflated the buyout by a fifth or more, and the support floor was then set against that inflated base for years afterwards. A discount on a bad number is still a bad number.

This is what separates a PULA exit from a ULA exit. At a ULA expiry the certification event arrives whether you are ready or not, which at least forces the count onto the calendar. A perpetual agreement supplies no such date, so nothing external ever forces the estate to be measured, and the first serious measurement tends to happen in the middle of a negotiation, under time pressure, using whatever the asset register happens to say. The register is almost always generous. It records what was bought and installed, not what is running today, and every stale record becomes a permanent line in the settlement.

The second finding is about framing rather than arithmetic. Exits opened on a corporate event closed 25 to 40 percent below exits opened on a renewal cycle. A renewal cycle is not a deadline on a perpetual agreement, because there is nothing to renew: the annual invoice arrives, you pay it, and the agreement continues. A sale agreement is different. It carries a completion date set by a transaction rather than by a licensing conversation, and it obliges both sides to resolve entity coverage before that date. That is the only clock in the relationship that does not belong to Oracle, which is why it is worth raising nine months before completion and worth very little four weeks before.

The third finding is the one buyers plan for least. Audit requests landed within twelve months of exit in roughly four of every ten cases we tracked. This is not retaliation; it is the predictable consequence of the cover disappearing. During the agreement, unlimited deployment of the named programs made measurement uninteresting. After signature you hold a fixed count, the entity is still in scope, and every deployment above the certified number is ordinary unlicensed use. Run a mock audit at month six and again at month nine, keep the written record of every decommission, and treat the twelve months after signature as part of the exit rather than as the period after it. The wider instrument is described on the Oracle PULA page, the instrument choice in the PULA versus ULA pillar, and the rest of the library in the Oracle practice.

Watch the briefing · 4:30How to Negotiate an Oracle ULA: No Price List, Just Your Business CaseThere is no price list: the fee is a story built from your estate and your growth, so give conservative answers and keep the product list narrow.

Across roughly 30 to 40 benchmarked Oracle PULA exits between 2024 and 2025, two figures separate the good outcomes from the expensive ones:

The patterns: settlements set against a register rather than a measurement, exits opened without a deadline anyone had to respect, and audit defense treated as a separate project rather than as the closing phase of the exit.

The buyer side move is to spend the first three months measuring and the next three negotiating. The wider library sits in the Oracle practice.

Can you exit a PULA mid term?

Only by negotiation. A perpetual agreement has no certification event and no expiry, so there is no moment that forces Oracle to the table. Every exit is a fresh contract, which means the leverage has to be built in the months before the first quote rather than found in the room on the day.

What are the four realistic exit paths?

A buyout that converts the perpetual right into a fixed counted entitlement, an assignment at a corporate event such as a divestiture or merger, a partial unwind that reduces the program list and the support fee together, and a migration linked step down where support falls against agreed migration milestones.

Why does the deployment baseline decide the settlement?

Because the number you certify becomes the number you pay on, for years. Across the exits we benchmarked, phantom seats inflated the opening buyout quote 20 to 35 percent, and the median overstatement was 22 percent. The support floor is then set against that inflated base, so a weak baseline costs more than a weak negotiation.

Does the perpetual right transfer to a divested entity?

Not automatically. Assignment requires Oracle consent in most agreements, so the language has to be negotiated at the divestiture rather than afterwards. Raised before the sale agreement is signed it carries real value. Raised after, you have a deadline and no leverage.

How long does a PULA exit take?

Six to twelve months from baseline to signature, plus the audit defense year that follows. Compressing it is what produces the weak baseline, because the count is the one input that cannot be rushed and cannot be corrected once it is in a signed contract.

Will Oracle audit after the exit?

Frequently. Audit requests landed within twelve months of exit in roughly four of every ten cases we tracked. The unlimited cover is gone, the support stream has changed, and the entity is still in scope, so run a mock audit at month six and again at month nine and close the gaps before any request arrives.

What is the most expensive mistake in a PULA exit?

Opening the conversation before the deployment baseline is clean. Phantom seats hand the seller the only fact it lacks, inflate the settlement, and set a support floor you pay against for years afterwards. Baseline first, then let Oracle quote.