The VMware squeeze

When Broadcom bought VMware in late 2023, AT&T said its renewal quote went up by 1,050%. Tesco was offered $23.5m for a single year of support and is now moving 40,000 workloads and suing Broadcom for more than £100m.

Those stories made headlines because the companies are enormous. But what about other sized businesses? If this is what the squeeze looks like at the top of the market, it’s worth asking what it looks like for everyone else.

This is the first in a series of posts about moving away from VMware. We’re going to be straight with you throughout, including about the parts that are hard, because there is already plenty of sales material in this market and not much candour.

What has actually changed

Since the acquisition, Broadcom has ended perpetual licences, bundled individual products into larger subscription packages and moved pricing to a per-core model. Gartner puts the typical cost increase at 300 to 400%. European cloud providers have reported increases of 800 to 1,500% to the European Commission, and in March 2026, the industry body CISPE filed a formal competition complaint asking Brussels for emergency measures.

Broadcom’s own partners have had it harder still. In January, it told hundreds of European cloud service providers that their VMware partner contracts would not be renewed, with weeks to wind down. The enterprises running on those providers now sit mid-contract on platforms with an expiry date.

There are two dates that belong in every IT leader’s diary. First is 12^th January 2027, when the EU Data Act bans charges for switching cloud providers, egress fees included, and the second is October 2027, where vSphere 8 and VCF 8 will reach the end of general support, with the versions that follow being subscription only. So while the legal cost of leaving is about to fall, the commercial cost of staying is heading the other way.

According to senior analysts, leaving VMware is hard

Sales pitches often skip this when talking about VMware migration. Gartner analyst Paul Delory told The Register in May that customers who quit often end up with more complex, less capable infrastructure: there is no one-for-one replacement, alternative hyperscalers typically run fewer virtual machines per server so you buy more hardware, and you can end up managing several products where you used to manage one.

He describes real migrations, and the numbers back him up. CloudBolt surveyed 302 IT decision-makers in January and found 86% actively reducing their VMware footprint, but only 2 to 4% fully migrated. Nearly everyone wants to move, but almost nobody has finished. 

The difficulty of leaving is a critical part of the equation, but the half that this never prices in is what staying costs you. Money, obviously. But also, and importantly, control. 

Stay put and your roadmap, pricing and product bundle are decided each year by a vendor that has shown, publicly and repeatedly, how it treats customers that have nowhere else to go. Waiting does not improve those terms. It compounds them.

If you’re not AT&T, what happens?

There’s a strange comfort in the cases that make the headlines. At least the giants can fight back in court. A mid-size business cannot.

That said, mid-size organisations hold one real advantage the giants would envy. Tesco’s migration covers 40,000 workloads. Yours might cover 400. Smaller estates carry less legacy, fewer tangled dependencies and fewer of the 10-year-old systems nobody dares touch, which is why a mid-size migration is typically measured in months, while an enterprise one runs years. The most expensive option for most mid-size businesses is the one that often feels safest: do nothing, and pay whatever the next renewal says. That is still a decision. It just gets made by default, and invoiced annually.

What a sensible move actually looks like

A sensible path starts with discovery, because migration is a discovery problem before it is a technology one. Most estates contain workloads that nobody has mapped and dependencies that nobody wrote down, and every migration that goes wrong, goes wrong there first.

Once the discovery phase is complete, we always advise to start small, and start new. This could be putting a greenfield project on a second platform, leaving your most sensitive workloads exactly where they are or running the new environment properly, testing what breaks, and moving more only when the evidence says you should. Prepare, test, move in phases. The reassurance worth having is that this way, the risk can be planned and reduced. That is a very different promise from the process being “simple”.

Ilkari provides more than a destination for organisations planning to move away from VMware. Our sovereign infrastructure and cloud services offer the right platform, resilience, and operational foundations needed to support your workloads beyond the initial transition.

To help plan for a successful move, we support customers and integration partners to understand and navigate their target environments. We provide the infrastructure and platform support to make a confident transition. Migration execution, including discovery, testing, and cutover, remains with your internal teams or integration partner of choice. 

Next in this series, we’ll be exploring the questions to answer before you move a single workload, starting with the ones that derail migrations when nobody asks them early enough.

In the meantime, if your VMware renewal lands in 2027, the time to start counting down and doing something decisive is now.