The most powerful buyer side move on Oracle Java is to carve your dependence down to the few workloads that truly need it. Once the broad dependence is gone, the per employee metric loses its grip, and what remains is a small residual you can defend cheaply or carry while you finish the job.
Oracle's Universal Subscription is priced on your total employee count. Since January 2023 the metric counts every full time and part time employee, every contractor, and every temporary worker, regardless of whether they ever touch Java. That means the cost is set by your organization, not by how much Oracle Java you actually run. A company with a single Oracle Java workload and a company that runs it everywhere can face the same employee based price. The implication is liberating once you see it: the path to a smaller bill is not to use Oracle Java more efficiently, it is to need it in fewer and fewer places until the dependence is small enough to defend or remove.
Carving is the deliberate act of shrinking that dependence. You are not trying to optimize Oracle Java. You are trying to make it irrelevant to all but a handful of workloads, so that when you sit across from Oracle the conversation is about a narrow residual rather than your entire headcount. The smaller and clearer that residual, the weaker Oracle's position and the stronger yours.
You cannot carve what you cannot see. The first task is a complete inventory of where Oracle Java genuinely runs, as opposed to where Java of some kind runs. Many estates discover that a large share of their Java is already a free OpenJDK distribution shipped inside a platform or a managed service, carrying no Oracle dependence at all. Others find Oracle runtimes hidden inside packaged software, build agents, and old desktop installs. The goal of the inventory is to separate three things: workloads that already run free Java, workloads that run Oracle Java but could move, and workloads that run Oracle Java because of a real constraint.
The size of your residual is decided by how hard you challenge the reasons workloads claim to need Oracle Java. Most reasons do not survive scrutiny. A vendor that only certified Oracle Java three years ago may certify a free OpenJDK distribution today. A feature that once required a commercial runtime may have a free equivalent in your current version. A workload pinned to an old Java release may simply need an upgrade that was overdue anyway. For each workload in the hold bucket, ask what specifically breaks if it runs on free Java, and require evidence rather than assumption. The residual that survives an honest challenge is usually a small fraction of the residual you started with.
Consider an anonymized manufacturer that believed it needed Oracle Java broadly across its estate. The inventory showed that more than half of its Java was already a free distribution bundled in platforms and tools. Of the genuine Oracle runtimes, most carried no real constraint and moved to a free OpenJDK distribution in two phases. What remained was a small set of workloads behind one engineering application that certified only Oracle Java. The firm carried that narrow residual on a tightly scoped agreement while it pressed the vendor for OpenJDK certification, which arrived within a year and let it remove the last dependence. The exposure tied to the broad employee metric fell to a small fraction of the opening number. The figures are indicative, but the method, see everything, move the movable, isolate and challenge the rest, is the one that works.
| Stage | What you are left with |
|---|---|
| Before carving | Oracle Java assumed across the estate, priced on full headcount |
| After discovery | Clear view of free Java versus genuine Oracle dependence |
| After migration | Only constrained workloads still on Oracle Java |
| After challenge | A small residual scoped to real, current constraints |
Once the residual is small and documented, the commercial work begins. Oracle will want to price even a tiny residual as if it justified a subscription across your whole employee count, because the metric is written that way. Your defense is the inventory and the migration record. They show that the dependence is confined to a named set of workloads, that the rest of the estate has moved, and that you have a credible plan to remove the last constraint. With that in hand you negotiate the residual against the smallest defensible envelope rather than your full headcount, which is where the largest savings against the opening position come from.
A carved estate drifts back to Oracle Java if nobody guards it. The free OpenJDK distributions you standardized on must be the default that new workloads inherit, and pulling Oracle Java back in should require a documented exception with a named owner. Govern the runtime the way you govern any platform standard, and review the residual on a schedule so that a constraint you resolved does not quietly leave a workload sitting on Oracle Java out of habit. The carve is only as durable as the governance that follows it.
A small, documented residual is also your strongest audit defense. LMS audits intensified in 2026 with a three year lookback that probes where Oracle Java has run and how many employees it should be counted against. A carved estate answers that probe cleanly. The inventory shows which workloads ever used Oracle Java, the migration record shows when each moved off it, and the residual shows the narrow set that remains. Instead of an open question priced against your full headcount, the auditor sees a confined, evidenced footprint. The carve down and the audit defense are the same body of work, which is why doing it before an audit lands is far better than doing it under one.
Order the carve to capture value early. Begin with discovery, because it often reveals that a large share of your Java is already free and that your true Oracle dependence is smaller than feared. Move the unconstrained workloads next, since they remove the most exposure for the least effort and prove the pattern. Then work the constrained residual down case by case, treating each as a small project to certify the free runtime, upgrade past the issue, or retire the workload. Carrying the residual on a tightly scoped agreement while you finish is usually cheaper than waiting to negotiate until every last workload has moved. The sequence captures the bulk of the saving long before the final workload is done.
Keep the residual visible to the people who decide budgets. A short, current summary of what remains on Oracle Java, why, and the plan to remove it keeps the carve from stalling once the easy wins are banked. The last workloads are often the ones with the most inertia, and a standing record of the residual is what keeps attention and funding on finishing the job rather than letting a small, expensive dependence quietly become permanent.
The buyer side goal is a small, documented residual. See every place Oracle Java actually runs, move everything without a real constraint to free OpenJDK, challenge every reason to stay, and price what remains on a narrow envelope. A small residual is a weak position for Oracle and a strong one for you.
Carving the residual is the destination of the migration strategy. For the full method see the OpenJDK Migration Playbook. To understand the staged approach that gets you here without a risky cutover, read about the partial migration strategy for Java, and to turn a small residual into bargaining power, see how migration becomes renewal leverage.
We can run the inventory, challenge the dependence, and help you price the residual on the narrowest defensible envelope so your Oracle Java exposure ends up where it belongs. Book a strategy call and we will scope the carve down for your estate.
We map your estate, isolate what truly needs Oracle Java, and build the migration that shrinks your employee envelope. Fixed Fee from 18,000 dollars or Gainshare with no risk to you.
Book a Strategy Call Get a QuoteFixed fee or gainshare, both backed by our guarantee. We sit between you and Oracle and we never take vendor money.
Get a QuoteWeekly intelligence on Oracle Java licensing moves and the buyer side defenses that work.