Certifying out of a ULA is a board decision, and a board needs a model, not a recommendation. Here is the cost case that compares certifying against renewing over the same horizon, the assumptions to test, and the numbers that actually move the answer.
The exit business case sets two paths side by side over the same multi year horizon: certify and keep a perpetual count, or renew the unlimited right. Because certification is free and a higher count costs nothing in support, the comparison usually turns on renewal fees, ongoing support, and migration costs, not on the certification itself. Model it honestly and the number speaks for itself.
A credible case has four moving parts. The renewal cost is what Oracle would charge to extend the unlimited right for another term, remembering that the first quote is an opening position and typically moves 20 to 40 percent. The certified value is the perpetual count you could defend at exit, which often lands well above a first estimate once cloud, disaster recovery, and non production deployments are counted properly. The support line continues under either path at the ULA level, so it is broadly common to both and rarely the deciding factor. And the migration cost captures any cloud move or platform change the exit path requires. Every figure here is estate specific, so the model is built from your numbers, not from benchmarks.
The certified count is the asset most exit cases undervalue. It is free, permanent, and yours, and it is usually the largest single number in the model.
Most weak business cases fail on the same four assumptions, and each is testable. The first is undercounting the certifiable deployment, which understates the value of certifying and tilts the case toward renewal by default. The second is treating Oracle's first renewal quote as fixed, when in practice these quotes move materially under pressure. The third is ignoring the carrying cost of shelfware, the support you pay on perpetual licenses you never deploy. The fourth is assuming a migration completes on time and on budget, when migrations routinely slip. Test each, and the case becomes something a board can sign rather than a position someone is advocating.
The figures below are indicative and exist only to show how the model behaves. They are not a quote and not a benchmark.
| Line | Certify and exit | Renew the ULA |
|---|---|---|
| Renewal fee, next term | None | Opening quote, movable 20 to 40 percent |
| Perpetual count secured | Full defensible count, free | Unlimited right continues, nothing locked in |
| Support, next term | Continues at ULA level | Continues at ULA level |
| Growth coverage | New licenses bought as needed | Covered by the unlimited right |
| Best when | Deployment is stable or shrinking | Heavy, certain growth ahead |
The pattern the table shows is the one most estates follow. If growth is stable or declining, certifying captures a free perpetual asset and ends the renewal fee. If you are certain of heavy growth that only the unlimited right would cover economically, renewal earns its place. The honest model tells you which world you are in.
A business case is only as strong as the count behind it, which is why the first move is an independent measurement of your defensible deployment. With a real certified value in hand, the comparison against renewal is straightforward and defensible to a board. Two companion notes go deeper: shrinking the estate versus growing the count on the strategic choice, and keeping leverage when Oracle pushes OCI on handling the cloud pitch that often arrives with the renewal. The wider method sits in our ULA exit strategy guide.
The renewal cost over the next term, the value of the perpetual count you would certify, the ongoing support either path carries, and the cost of any migration or cloud move. Set the two paths side by side over the same horizon. The certified count is free value, so the comparison usually turns on support and renewal fees, not on certification.
Often, but not always. Certifying ends the renewal fee and converts your deployment to perpetual licenses you keep, while support continues at the ULA level. Renewal makes sense mainly when you expect heavy growth that the unlimited right would cover cheaply. The honest model over a multi year horizon decides it for your estate.
Underestimating the certifiable count, treating Oracle's first renewal quote as fixed when quotes move 20 to 40 percent, ignoring the carrying cost of shelfware, and assuming a migration will complete on time. Test each assumption, because the case is only as good as the numbers behind it.
We measure your defensible deployment, model certifying against renewing over the same horizon, and hand you a case a board can sign. Book a confidential assessment.