Support, not the license fee, is the number that decides certify or renew over the long run. Certification holds support flat at the existing ULA base. Renewal adds a fresh fee and lifts a base that compounds for years. Model both over five to ten years before you choose.
By the Meridian advisory team, former Oracle LMS and GLAS licensing analysts. Updated 4 June 2026.
No. At certification, support continues at the existing ULA level regardless of how many licenses you certify, so certifying a higher count does not raise your support bill. The widespread fear that a larger certified number lifts support is a myth, and an expensive one, because it pushes organisations to under count and leave entitlement behind. Support only changes later, through Oracle's standard annual uplift or a separate repricing exercise, never because the certified number was high. That single fact reshapes the whole certify or renew model.
Certification is free and the certified count does not move support. Renewal buys time at the cost of a fresh fee and a higher support base that compounds. The license fee is visible; the compounding support is where the real money sits.
Model both paths over the same horizon, five to ten years, and compare cumulative cost, not the headline fee. Under certification, support carries forward from the ULA base and rises only by the annual uplift, while the license fee is zero. Under renewal, you pay a fresh license fee, and the support base is reset higher and then compounds by the uplift every year afterward. The renewal fee is a one time number that draws the eye, but the lifted support base is the recurring number that accumulates quietly. Over a decade, the compounding base often dwarfs the fee that anchored the conversation, which is why the headline quote is the wrong thing to optimise. The pitch built around that quote is dissected in how Oracle sells the renewal.
Take an indicative organisation with an existing ULA support base, expressed here as 100 units a year, and an Oracle style annual uplift of 4 percent. Under certification, the license fee is zero and support continues from 100, rising 4 percent a year. Under renewal, assume a fresh license fee of 250 units and a reset support base of 130 units, also rising 4 percent a year. The figures are indicative and chosen to illustrate the shape, not to predict any real quote. The point is the trajectory, not the precise numbers, which always depend on your contract.
| Cost element | Certify path | Renew path |
|---|---|---|
| Upfront license fee | 0 | 250 |
| Year 1 support | 100 | 130 |
| Annual uplift | 4% | 4% |
| Support over 5 years (approx) | 542 | 704 |
| Five year total (fee plus support) | 542 | 954 |
| Ten year total (approx) | 1,201 | 1,811 |
In this indicative model the renewal path costs more over both horizons, and the gap widens with time, because the higher support base compounds. The fresh fee of 250 is visible, but most of the difference comes from the lifted base. Reverse the conclusion only if real growth means the renewed unlimited right captures deployment worth more than the gap. That is the genuine case for renewing, and it has to be evidenced, not assumed.
If your deployment of the named products will genuinely expand fast, the renewed unlimited right lets you grow without buying licenses, and the value of that capacity can exceed the compounding support gap. The test is whether the growth is forecastable and defensible, not hoped for.
On the certify path, once you hold perpetual licenses you have more freedom to manage support, including retiring support on shelved licenses where the contract permits. That can bend the certify curve downward further, widening the advantage. The decision to shelve or keep is its own analysis.
The annual uplift drives the compounding on both paths, so the model is sensitive to it. A higher uplift makes the renewal base more expensive over time, strengthening the case to certify and hold the lower base. Run the model at your actual uplift, not a generic one.
The certify or renew decision is usually won or lost on support, not on the license fee, and the only honest way to see that is to model both paths over years, at your real base and your real uplift. Certification holds the base flat and costs nothing to execute; renewal lifts a base that compounds. Unless provable growth closes the gap, the long run number tends to favour certifying. The five inputs that frame the choice are in the five questions that decide certify or renew, and the full method lives in our pillar, the certify or renew guide.
Support is the decisive number. Certifying holds support flat at the existing base regardless of the certified count, while renewing adds a fresh fee and lifts a base that compounds every year. Model both over five to ten years at your actual uplift. The renewal fee is visible, but the compounding support is where the cost lives, and it usually favours certifying unless real growth closes the gap.
No. At certification, support continues at the existing ULA level regardless of how many licenses you certify, so certifying a higher count does not raise support. The fear that a larger certified number lifts support is a myth. Support only changes later, through Oracle's annual uplift or a separate repricing, not because of the count.
Model both paths over five to ten years. Certify holds support flat from the ULA base, with an annual uplift. Renew adds a fresh license fee, then lifts the support base that compounds each year afterward. Comparing the cumulative totals usually shows the renewal fee is the smaller part of the long run cost.
Book a confidential assessment and we will build the certify versus renew cost model at your actual support base and uplift, so the decision rests on evidence.