Full narration of the briefing. Click a section heading to jump the player to that moment.
An Oracle unlimited license agreement is unlike anything else Oracle sells, because there is no price list. None. The fee is constructed from a business case Oracle builds about you: your estate, your plans, and above all, your growth. Which means every ULA negotiation is really a negotiation about a story, and whoever controls the story controls the price.
Here are the five rules.
Rule one. Understand what you are actually negotiating. With no published pricing, the ULA fee is derived from what Oracle believes you would otherwise spend: your current deployment, your projected growth, and the license fees that growth would generate at list. Every discovery question, every architecture workshop, every friendly conversation with your engineers is input to that model.
Treat the entire pre-sales process as what it is: Oracle building the case for the largest defensible number. Your job, from the first meeting, is to control the inputs.
Rule two. When Oracle asks for your growth estimates, and they will ask, early and often, give conservative, defensible projections. Your optimism is their pricing model: every data center expansion you mention, every migration you speculate about, every headcount plan you share converts directly into the fee. Do not volunteer roadmaps in discovery.
Do not let enthusiastic architects brief Oracle unsupervised. State the growth you are contractually comfortable being priced on, grounded in approved budgets, and nothing more. You can always deploy more inside a ULA. You can never un-tell them your plans.
Rule three. Fight for a narrow product list. Oracle will propose a broad ULA: database options, middleware, tools, everything in the estate. Every product on that list inflates the fee, and worse, inflates the support stream that continues after the ULA ends, forever.
Include only the products where genuine, near-certain growth justifies unlimited deployment, typically a handful of high-growth, mission-critical items. Everything else stays outside on regular licenses. A narrow ULA is a tool. A broad ULA is a mortgage on software you never deployed.
Rule four. Run the breakeven before you believe the pitch. A ULA pays off only when your real deployment growth generates more license value than the fixed fee costs; flat estates lose, every time. Model it with your conservative growth number: the accumulated per-processor cost of what you will actually deploy, against the certain fee plus its support stream.
And weigh Oracle's favorite closing argument properly: the threat of the eight percent annual support increase is real, but capping support is negotiable on its own, without buying an unlimited agreement to escape it.
Rule five. Negotiate the certification before you sign, not before you leave. The ULA's real price is discovered at the end, when you certify your deployment and it becomes your perpetual license count. Fix the exit terms at entry: how cloud deployments count, what evidence certifies the number, that certification is not an audit event, and what the support base becomes afterward.
Then, nine to twelve months before term end, build one clean, defensible deployment baseline across license data, configuration, and hypervisors. Companies that certify from their own count keep what they deployed. Companies that certify from Oracle's count fund the next ULA.
One last point. At Redress Compliance we negotiate and certify Oracle ULAs for large enterprises on a pure contingency basis. Our fee is 25 percent of what we save you. If we save you nothing, you pay nothing.
Before you share a single growth number with Oracle, talk to us. com.
Redress Compliance works on contingency: our fee is 25 percent of what we save you. Nothing saved, nothing paid. Independent, buyer side only, never vendor funded.
Talk to a Oracle negotiator