Renewal is Oracle's preferred outcome, but for many estates certification is the obvious win. These are the signals that point clearly toward certifying, and how to read your own position against them.
By Daniel Voss · Ex Oracle LMS · 4 June 2026
Certifying is clearly right when your deployment has stabilized, your defensible count is high relative to a renewal fee, and you have no concrete plan to expand the named products materially next term. Certification carries no fee and converts deployment into permanent entitlement at the support level you already pay, so a mature estate usually captures more value by certifying a maximized count than by paying again for an unlimited right it will not use. The decision still turns on your specific contract language.
A ULA grants unlimited deployment of named Oracle products for a fixed term and fee. At the end you either certify, converting your deployed quantities into a permanent entitlement, or renew, paying again for another term of unlimited rights. Oracle presents renewal as the safe continuation, and the renewal quote arrives early and confidently. That framing is a commercial position, not a neutral recommendation. For a large share of estates, certification is the move that captures value, and renewal is the move that keeps paying for a right the organization has stopped using.
The task is to read your own situation honestly against a small set of signals, rather than accept the default. When several of these signals are present, certifying is not a close call. It is the clear choice, and the work shifts from whether to certify to how to maximize the count you take with you.
It is clearly right when the estate has matured and the unlimited right has stopped earning its fee. Concretely, look for these signals. The more of them you see, the stronger the case to certify.
Your Oracle footprint for the named products has flattened. You are not in a phase of rapid new deployment that an unlimited right would absorb cheaply. A stable estate has already captured most of the value a ULA can give, and the next term would largely pay for capacity you will not add.
Once cloud, disaster recovery, test, and virtualized deployments are counted properly, your defensible number is large relative to what renewal would cost. Certified counts often land well above a first rough estimate, frequently in the range of one and a half to two and a half times higher once the full estate is measured. That uplift is permanent entitlement captured for no fee.
You have no funded, scheduled plan to grow the named products materially in the next term. A genuine expansion can justify renewing, but a vague possibility of growth is not a reason to pay again. Deliberate future purchases can be made when the need is real, often more cheaply than a renewal that prices in growth that may never happen.
Renewal generally resets and often raises the support base. Certifying fixes your entitlement and lets you manage support against a known number, including terminating support on licenses you do not need. An estate focused on cost control usually prefers the fixed position to another open ended term.
Certification is free and permanent. Renewal is paid and temporary. When your estate has stopped growing into the unlimited right, the question is not whether to certify but how large and defensible a count you can take with you. Maximize the number, then walk away with it owned.
One belief sends more sound certification candidates into renewal than any other: the fear that certifying a larger count will raise support fees. It will not. Support continues at the level set by the ULA regardless of the certified number, so a higher count is free permanent value. Certifying the largest number you can defend with evidence increases your entitlement at no added support cost. If this fear is the main thing pulling you toward renewal, it is worth debunking explicitly before the decision is made, because it is simply not how the mechanics work.
An anonymized example shows the shape of the choice. A logistics group held a ULA on database and several options, with a footprint that had been flat for two years. A first estimate put the count at roughly 600 processors. A full measurement, including disaster recovery and a virtualized cluster, produced a defensible 1,300. The figures are indicative, but the logic is clear: certifying captured more than twice the entitlement of the rough estimate, at no fee and no support increase, against a renewal quote that would have paid again for capacity the group had no plan to add. With stable deployment, a high defensible count, and no expansion roadmap, certification was not a close decision.
If your deployment has stabilized, your defensible count is high, and you have no funded expansion plan, certifying is the clear move, and the work is to maximize the number you take. Read the framework end to end in our certify or renew guide, see how the choice becomes negotiating power in using the certify option as leverage, and understand the middle path in the hybrid outcome, certify some, renew some.
Book a ULA assessment and we will measure your defensible count, weigh it against a renewal, and tell you plainly whether certifying is the clear move for your estate.