Full narration of the briefing. Click a section heading to jump the player to that moment.
This is the episode where we are going to argue against ourselves, deliberately. If you walk into a Broadcom negotiation saying we will just leave, and you have not done the work, you will be taken apart, because their account team knows the completion statistics better than most customers do. So we are going to give you both halves. Where leaving genuinely saves money, and where it genuinely does not.
I am Tom, Daniel is with me, and the reason this matters is that a position you can defend under questioning is worth far more than a bolder one you cannot.
Start with the numbers that hurt. Only about four percent of enterprises have completed a full VMware replacement. Two percent are past seventy five percent migrated. Meanwhile eighty six percent are reducing their footprint, which tells you the direction of travel is real but the finish line is rare.
Gartner puts migration services at five hundred to a thousand dollars per virtual machine over eighteen to thirty six months, and says explicitly that cost savings alone should not be the driver. Now layer on the market: server list prices rose fifteen to thirty percent in 2026, and DRAM contract prices as much as ninety to ninety five percent in a single quarter.
And the alternatives have moved. Gartner's clients report alternative costs similar to Cloud Foundation, not dramatically below it. Nutanix's own filings show rising average selling prices and contract terms lengthening to three point four years, and its chief executive is on record that server supply constraints are pushing customers toward public cloud. Then there is the point almost nobody makes out loud.
Every credible alternative in this market is private equity owned and looking for an exit. BMC splitting under KKR. Rocket under Bain. Precisely on its third owner.
Leaving Broadcom changes your counterparty. It does not exit the asset class.
On the mainframe the answer is not one answer, it is a ranking, and knowing the ranking is what makes a threat credible instead of embarrassing. The security managers are genuinely replaceable. There is named conversion tooling that exists and is downloadable, and a documented ACF2 to RACF conversion completed in six months, deliberately timed to land before the Broadcom renewal. Endevor is replaceable on real evidence: two German banks completed removals in late 2025 with IBM Expert Labs, one moving a hundred and eighty one thousand elements in eleven months against an eighteen month plan.
But Datacom and IDMS have no replacement at all, only an application rewrite. Never bluff there.
The sharpest precedent in the whole file is a Dutch insurer, ASR. It eliminated all five CA products it ran. It moved roughly four thousand CA-Telon programs in nine months, using its own developers rather than a system integrator, and it retired the contract on its finish date rather than renewing and migrating afterwards. The IDMS to Db2 work took about a year and a half.
Total saving, roughly three million euros a year. That is the shape of a credible exit: sequenced against the contract end date, resourced internally, and started early enough that the renewal was never the forcing event. That is what you are aiming at.
Balance it with what the analysts say, because Broadcom will. Gartner's position, published in April 2026, is that more than seventy percent of mainframe exit projects started that year will fail to produce the intended benefits, that only around ten percent of mainframe users will migrate off by 2030, and that three quarters of the vendors selling mainframe exits will pivot or cease to exist by then. The technical shortcut closed too, when the English court found against LzLabs in March 2025 and the Court of Appeal refused permission that July. A platform exit is not believed.
Displacing named products at a renewal date is.
So where does switching actually save. Four places, consistently. Where the licensing metric changes, per socket or per node against per core, which is a structural saving rather than a discount. Where you already own Windows Server Datacenter with Software Assurance and are paying twice for virtualisation rights.
Where a small estate is caught by core minimums and is paying for capacity it does not have. And where you are refreshing hardware anyway, so the migration cost rides on a budget that already exists. And the best line in the room is Broadcom's own conversion literature, which argued exactly this when it was the challenger. Next time, where your leverage actually is.
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. Want Redress to contact you? Reach out and we respond the same day.
Talk to a Broadcom negotiator