Tesco just sued Broadcom for £100M+ (for us Yankees, about $132M). Forty thousand server workloads. A grocery retailer with its own data center empire, a deep bench of infrastructure engineers, and the legal budget to fight back. Good for them.

You are not Tesco.

If you’re running a mid-market tech org – call it 100 to 200 VMs – the Tesco story is instructive, but your licensing budget and decision paths are going to look very different.

The VMware Tax Is Real, and Mid-Market Pays Full Price

When Broadcom acquired VMware, they did what aggressive acquirers do: they monetized the moat. Perpetual licenses became subscriptions, and the bill went up – by most accounts, dramatically.

Enterprise shops took the hit and opened contract negotiations. They have leverage – volume commitments, multi-year relationships, dedicated account teams, and the real ability to exert legal pressure. The math is painful but likely survivable.

Mid-market doesn’t have those levers. You’re paying close to rack rate on licensing, running a lean infrastructure team, and provisioning hardware for peak load that sits underutilized the rest of the year. The VMware cost increase is closer to existential at your level, and it now demands a strategic answer.

Why Tesco Isn’t Going Cloud — And Why You Should Seriously Consider It

Tesco is moving to another on-prem platform, not Azure, AWS, or GCP. That decision makes sense for them and may not for you.

At 40,000 server workloads, you’re managing a data center estate. You have engineering staff whose entire career capital is on-prem infrastructure. Retraining that many people to think in cloud-native terms – infrastructure as code, auto-scaling, cloud networking primitives – is more than just a training problem. It’s a culture shift, and at Tesco’s scale I’d wager that’s a multi-year program.

Then there’s the data problem. You don’t just lift and shift 40,000 server workloads. So, this is likely a hybrid setup at best. A retailer the size of Tesco is running transaction volumes that would push enormous amounts of data between the cloud and on-prem infrastructure nodes. This means egress charges – and at Tesco’s scale, the math on that will tilt heavily in favor of keeping things on-prem.

Your mid-market situation is different on every dimension. Smaller team. Less sunk cost. No data center empire to protect. And critically – you’re a tech org in service of a business that likely sells something else. Every dollar you spend on rack space, hardware refresh cycles, and VMware licensing is a dollar that didn’t go toward building the product or serving the customer.

None of this, however, means you should lift 150 always-on VMs into Azure and have a victory parade.

It does mean Broadcom just gave you a reason to evaluate your estate workload by workload: what should be retired, what should remain on-prem, and what would benefit from cloud elasticity?

In my experience, the cloud is the right fit for most mid-market workloads. Not all. But most.

What Goes Up Can Actually Come Down

On-prem infrastructure gives you limited ability to scale down. You can buy more servers. You cannot un-buy them.

Mid-market companies almost always provision on-prem infrastructure for peak load – the quarter-end crunch, the product launch, the seasonal spike. The rest of the time, that capacity sits idle.

You paid for it. It’s yours. Congratulations, I guess.

Cloud flips that model. For workloads that can scale dynamically – or be shut down when they are not needed – you pay for what you use, scale for what you need, and release it when you don’t.

That flexibility is not free. A poorly planned lift-and-shift can turn server depreciation into an even larger monthly bill.

But for a company expecting to grow – or one navigating the uncertainty of not knowing exactly how fast – paying for elasticity can be more valuable than owning idle capacity. It protects the P&L from large upfront bets and allows infrastructure to follow demand on a near just-in-time basis.

Tesco Called Broadcom’s Bluff. You Can, Too.

Broadcom bet that enterprises would absorb the price increase rather than pay the switching cost. That bet likely pays out highest for enterprises at large scale, where switching is the most painful – but, as Tesco shows, most of the time is not all of the time.

For mid-market, Broadcom may have inadvertently done you a favor – making VMware economics untenable enough to force an infrastructure conversation you probably needed to have anyway.

The right answer for you may be cloud, another virtualization platform, or a deliberate mix of both. Regardless, staying put is no longer the automatically conservative choice. Not when you’re staring at a 175% increase in a material budget line item.

I’m on record favoring the cloud, though I’m pragmatic enough to know the cloud isn’t right for every workload.

But if you’re trying to explain an infrastructure budget to a CFO who wants to know why you’re spending capital on refreshing servers every 3-5 years instead of product – the VMware licensing change just handed you the argument you needed.

Use it.