For two decades, virtualisation was the quiet constant of enterprise IT. It was rarely questioned because it rarely needed to be. That has changed. Licensing restructures, bundled subscriptions and rising renewal costs across the VMware ecosystem have forced many organisations to confront a platform they had long treated as settled.

The immediate pressure is financial, but the deeper issue is strategic. When one vendor controls the layer beneath every application, your pricing, roadmap and risk posture are partly set by someone else. Many leaders are discovering that their estate is simply inherited, not deliberately designed.

The cost of standing still

The instinct is to renegotiate and carry on. That buys time but not options. Meanwhile, the demands on infrastructure are shifting. Enterprises are preparing for AI workloads, tightening data sovereignty requirements and managing applications across several environments. The estate was not built for any of this.

Hybrid is now the norm. More than 80% of organisations believe hybrid IT is the best way to manage applications and data, and 81% plan to sustain or expand their hybrid approach over the next three years. The question is no longer whether to spread workloads across environments. It is whether your virtualisation layer helps or hinders that.

Three paths, not one

Disruption is easier to manage when the choices are clear. In practice, organisations tend to take one of three routes:

• Re-platform to virtual servers in the cloud. A like-for-like landing zone gives a fast exit from legacy tooling, and early benchmarks suggest total cost of ownership more than 20% lower.
• Re-platform onto a container-native stack. Running virtual machines alongside containers on OpenShift Virtualization avoids refactoring and keeps control with your own teams, with savings of 15% or more.
• Re-factor for the long term. For applications worth reinvesting in, a hybrid-by-design approach can deliver up to three times the return.

These are not competing ideologies. Most estates will use all three, matched to the workload.

Migration as a launchpad

The strongest migrations treat the move as a platform decision, not a rescue operation. Done well, the same stack that hosts today’s virtual machines can run tomorrow’s containers and AI services, with consistent security, governance and automation. Organisations taking this approach report operating cost reductions of around 35%, and they gain regulatory confidence along the way: encryption, data residency and jurisdictional control that can be demonstrated, not just promised.

Where to begin

Start with an honest inventory. Which workloads are tightly coupled, which are candidates for modernisation, and which carry sovereignty obligations? Then test assumptions with a pilot rather than a full commitment.

IBM has been working with organisations through exactly this process, from Sydney’s regional footprint to hybrid environments spanning Power, x86 and OpenShift. If your next renewal has prompted these questions, it’s worth exploring what an open, lower-lock-in path could look like.