Why EOL Events Are a Procurement Problem
An end-of-life event is a commercial event wearing a technical disguise. When a vendor ends support for a product, it manufactures a deadline that converts switching costs into leverage — and the enterprises that pay the most are the ones that treat it as an IT upgrade rather than a procurement negotiation. Managing EOL well is part of the risk discipline set out in the IT procurement transformation guide: a function with visibility and lead time has options, while a reactive one has only the vendor's deadline.
The Broadcom-VMware Cautionary Tale
The defining recent case is Broadcom's handling of VMware. After the acquisition, Broadcom moved customers off perpetually-licensed virtualisation onto VMware Cloud Foundation subscription bundles, raising costs sharply for many and bundling in products they did not use. vSphere 7 reached end of general support on 2 October 2025, and an October 2025 partner restructuring cut the VMware partner ecosystem from roughly 4,500 to around 200 authorised providers. Tesco's litigation — alleging it would take at least three years to migrate off Broadcom's VMware and mainframe products — shows exactly how a vendor turns switching cost into pricing power. The full playbook for this specific transition is in the VMware-Broadcom Survival Guide, and the VMware/Broadcom vendor hub tracks the licensing changes.
The Three Strategic Options
A disciplined function models three options rather than accepting the default. Upgrade to the vendor's successor product or subscription bundle — the path the vendor wants, often the most expensive. Extend the runway with third-party support while delaying the upgrade, buying time and negotiating leverage. Migrate to an alternative platform, the highest one-off cost but the route that ends the dependency. The error is letting the EOL deadline force the upgrade by default before the other two are even costed.
| Option | Best when | Key risk |
|---|---|---|
| Upgrade to successor | Switching cost high, repricing moderate | Locks in vendor's new commercial model |
| Third-party support | Need runway; upgrade not yet justified | No new features; vendor relationship strain |
| Migrate to alternative | Repricing aggressive; viable alternative exists | High one-off cost; multi-year timeline |
The VMware episode is instructive precisely because it was not unique — it was a template. The same pattern recurs whenever a vendor is acquired by a portfolio owner or shifts from perpetual licences to subscription: support is narrowed, bundles are widened, and the price of staying climbs steeply for customers whose switching costs the vendor knows are high. Oracle's Java repricing and IBM's metric enforcement follow the same logic. Treating each as a one-off surprise is the mistake; a mature function reads the acquisition or the licensing-model change as the early signal it is, and starts modelling alternatives while the deadline is still distant enough to preserve every option.
Building the TCO Models
The decision turns on total cost of ownership, not headline price, modelled over a multi-year horizon. Each option carries different cost shapes: upgrade front-loads subscription increases, third-party support trades features for time, and migration front-loads project cost against lower steady-state spend. The model should capture implementation, integration, retraining, and exit costs — the same TCO discipline that underpins the procurement KPI framework. Comparing the three on a like-for-like, risk-adjusted basis is what converts an emotional "we can't move" into a defensible commercial choice.
The horizon matters as much as the cost shapes. A three-year TCO will favour the upgrade because the migration project cost lands up front, while a five-year view often flips the answer as the avoided subscription increases compound. Model both, and make the time horizon an explicit decision rather than an accident of whichever spreadsheet someone built first. Build in a sensitivity for the vendor's likely next price move, too: an incumbent that has just repriced once will reprice again, so the "do nothing and upgrade" line should assume further increases rather than today's quote held flat. The TCO model is only as honest as the assumptions it makes about a vendor that has already shown its hand.
Negotiating From the EOL Position
Counter-intuitively, an EOL event is a negotiation opportunity if you arrive prepared. A credible, costed migration alternative is the strongest leverage you can hold against an incumbent's repricing — the vendor's deadline cuts both ways once you can walk. This is where the vendor scorecard and a clear-eyed read of the relationship-versus-management distinction pay off: an incumbent that has underperformed has weakened its own hand. For multi-vendor estates, the Multi-Vendor Strategy white paper sets out how to keep a credible alternative warm so the threat is real, not rhetorical.
Leverage decays the moment the deadline passes, so timing the negotiation is itself a tactic. The strongest position is held a year or more before end of support, when migration is still a credible threat and the vendor still wants the renewal signed cleanly; the weakest is the week before the deadline, when extended support or an emergency upgrade is the only option left. Vendors understand this asymmetry perfectly and pace their concessions to it, which is why a function that engages early extracts terms a function that waits never sees. The entitlement-accurate licence inventory is what lets you move early with confidence, because you cannot credibly threaten to migrate an estate you cannot fully describe.
An EOL Readiness Playbook
Readiness beats reaction. Begin modelling 18–24 months before any end-of-support date, because infrastructure migrations run into years and third-party support takes months to arrange. Flag every product's lifecycle dates in the procurement calendar so EOL is anticipated, keep an entitlement-accurate licence inventory so you know exactly what you run, and cost all three options before the vendor's deadline collapses them to one. Where the stakes justify outside help, our software licensing negotiation practice models the decision and runs the transition — request a confidential briefing before the EOL clock starts setting your terms.