The renewal quote lands and it is three to five times what you paid VMware three years ago — in the worst cases advisors have documented, closer to ten. Under Broadcom’s subscription model that is the standard opening position, not a clerical error, and it is designed to be signed by teams that start the conversation ninety days out. Teams that start twelve months out tell a different story: licensing advisors consistently report disciplined negotiations removing 25–40% from the opening VMware Cloud Foundation quote, and infrastructure rightsizing alone typically cuts licensed cores 15–30% before commercial talks even begin.
This is the playbook — month by month, artifact by artifact. It assumes you already understand the mechanics of the new model (if not, start with our breakdown of the 2026 VMware licensing changes) and that you have not yet decided whether you are staying, leaving, or splitting the estate. Good. Undecided is the strongest negotiating position you will ever hold.
Why this renewal is different
Broadcom did not just raise prices. It changed the unit of sale. Perpetual licenses are gone, the à-la-carte catalog collapsed into a handful of subscription bundles centered on VMware Cloud Foundation, and everything is metered per core with a 16-core minimum per physical CPU — an 8-core or 12-core socket still bills as sixteen. Reports through 2025 also described a 72-core minimum order threshold in channel terms, with some flexibility at renewal for smaller sites; if you run edge locations, get that floor in writing before you assume it applies to you.
To be fair to Broadcom’s pitch: for shops that were already buying vSphere, vSAN, NSX and Aria separately, VCF per-core list pricing can land below the old sum of parts, and VCF 9 is a genuinely more integrated private-cloud stack than the catalog it replaced. The problem is the customers who were not buying all of that — they are now paying for a platform they use a third of. That asymmetry is why two enterprises with identical host counts can see renewal outcomes that differ by 3x. The takeaway for your steering committee: this is not a price increase to absorb, it is a repricing event to negotiate.
The 12-month timeline
Leverage decays with time. Every month you wait, an alternative becomes less credible because you can no longer execute it before your entitlements lapse. Work backward from the renewal date:
| What you do | Artifact you produce | |
|---|---|---|
| T-12 to T-9 months | Full inventory and rightsizing pass. Decommission zombie VMs, consolidate hosts, kill unused components. | RVTools export, per-host core map, target core count 15–30% below today’s. |
| T-9 to T-6 | Evaluate alternatives seriously. Run a Nutanix or OpenShift proof of concept on one production workload class. | A costed landing zone: hardware, licenses, migration labor, dates. |
| T-6 to T-4 | First commercial contact. Request a quote for your rightsized core count and only the tier you use. | Opening quote — treat it as a starting bid, not a price. |
| T-4 to T-2 | Negotiate term length, discount, and contract protections. Brief the CFO on the walk-away number. | Executive-approved BATNA and budget ceiling. |
| T-2 to T-0 | Final terms. Legal reviews audit clauses, true-up mechanics, renewal caps. | Signature — or a funded migration program. |
If you are inside six months right now, compress the first two phases into one and accept that your discount range narrows. Inside three months, your realistic play is a short-term renewal on the smallest footprint you can defend — then run this playbook properly for the next cycle.
Run your own core inventory before Broadcom does
The single cheapest source of savings is not negotiation — it is arithmetic. Broadcom’s quote will be built from your deployed footprint as its tooling and your account team see it. Your counter must be built from what you actually need. RVTools remains the fastest way to get there: export every host, socket, and core, then map physical CPUs against the 16-core minimum. A cluster of dual 12-core hosts is billed at 32 cores per host, not 24 — so on your next hardware refresh, fewer sockets with denser cores is the licensing-rational shape. Hold every line of the quote to actual CPUs and this rule, because over-counting at renewal is common and rarely in your favor.
Then rightsize. In twenty years around enterprise virtualization I have never seen an estate that could not shed 15–30% of its licensed cores once someone was paid to look: powered-off VMs holding reservations, dev clusters sized for a project that shipped in 2023, DR capacity licensed identically to production when a lower tier would do. Every core you remove before the quote is issued saves you the subscription price times the term length. Do this before first commercial contact, not after — a quote, once issued, anchors the negotiation.
The costed Nutanix landing zone — your most leveraged artifact
Saying “we’re looking at alternatives” moves a Broadcom quote by approximately nothing. Account teams hear it hourly. What moves the number is a document: a costed, dated landing zone for one production workload class on a competing platform — hardware quoted, licenses quoted, migration labor estimated, a named executive owner. You do not need to commit to leaving. You need to prove you can.
Nutanix is the most common vehicle for this, and for defensible reasons: its AHV hypervisor ships as part of the Nutanix Cloud Platform at no separate hypervisor cost, its Move tool handles VM conversion from vSphere with built-in replication and cutover scheduling, and — significant for the many shops already running ESXi on Nutanix hardware — its in-place cluster conversion can flip those clusters to AHV without swing hardware. The honest counterweights: a move usually pairs with a hardware refresh, so the capital timing has to line up; your operations team must retrain off vCenter muscle memory; and some deep VMware-ecosystem integrations (niche backup workflows, specific network and VDI tooling) need re-validation on AHV. Budget real dollars for that in the landing-zone cost model — an implausibly cheap exit plan is worse than none, because Broadcom’s negotiators can read a spreadsheet too. Our Nutanix vs. VMware cost analysis walks through the full five-year math, and OpenShift Virtualization fills the same role if your estate is heading toward containers anyway.
One more discipline: negotiate Nutanix like you negotiate Broadcom. Vendors who know they are the only exit route price accordingly. Two costed alternatives beat one.
The commercial negotiation itself
With a rightsized core count and a credible exit on the table, the 25–40% range advisors report becomes achievable. Where the concessions actually come from, as of mid-2026: term length (three-year commitments price materially better than one-year), tier discipline (refuse the VCF bundle if vSphere Foundation genuinely covers you — downgrading tiers is often worth more than discounting the big bundle), and timing (Broadcom runs on quarters like every enterprise vendor; a deal that can close inside its fiscal quarter finds flexibility a mid-quarter deal does not).
What Broadcom will rarely concede: the per-core model itself, the 16-core minimum, and support for a shrink-anytime clause. So get your protections in the contract instead — a renewal price cap, the right to reduce cores at renewal without penalty, and clearly defined true-up mechanics. A 35% discount on year one means little if year four reprices at list. Short version for the meeting: discount is the headline, contract language is the money.
Audit and true-up exposure is rising — pre-empt it
The subscription model gives Broadcom continuous visibility incentives the perpetual world never had, and audit and compliance-review activity has visibly increased since the transition. The exposed positions are predictable: hosts running entitlements from a lapsed contract, usage above subscribed core counts, and orphaned components from the old catalog still deployed. Anyone still on perpetual vSphere 8 should also note the clock — vSphere 8 support ends in 2027, which is precisely the pressure point an audit letter likes to arrive on.
The defense is the same core inventory you built at T-12. If you can produce an accurate, timestamped map of deployment versus entitlement, an audit becomes a reconciliation exercise instead of a settlement negotiation. If you cannot, assume the auditor’s count — and the auditor’s count includes the cluster someone stood up in 2021 and forgot. Self-audit before renewal, every cycle, without exception.
Frequently asked questions
How much can you negotiate off a Broadcom VMware renewal?
Licensing advisors report 25–40% off the opening VCF quote for disciplined negotiations backed by an accurate core inventory and a costed alternative. Rightsizing the estate first typically removes a further 15–30% of licensed cores, which compounds with the discount.
When should I start negotiating my VMware renewal?
Nine to twelve months before the renewal date. The inventory and rightsizing work takes a quarter, a credible alternative evaluation takes another, and your leverage erodes steadily as the lapse date approaches. Inside three months, aim for a short-term bridge renewal and reset for the next cycle.
What is the 16-core minimum in VMware licensing?
Broadcom licenses VMware subscriptions per physical core with a floor of 16 cores per CPU — a socket with 8 or 12 cores still bills as 16. Verify every quoted line against actual installed CPUs, and favor fewer, denser sockets at your next hardware refresh.
Can I still buy perpetual VMware licenses?
No. Broadcom ended perpetual sales and moved the portfolio to subscription bundles built around VMware Cloud Foundation and vSphere Foundation. Existing perpetual licenses keep running, but support renewal paths push toward subscription, and vSphere 8’s 2027 end of support puts a hard date on the holdout strategy.
Does threatening to move to Nutanix actually lower the quote?
A verbal threat, no. A costed landing zone with quoted hardware, quoted Nutanix licensing, migration labor estimates, and executive backing, yes — it is the single most leveraged artifact in the negotiation, because it converts “we might leave” into a number Broadcom must price against.
Enterprise Techie publishes vendor-honest analysis like this daily — get the brief by email, free.