Sponsored by Broadcom
To some executives, mainframes are the most “mysterious” and scrutinized line item, sitting at the top of the IT budget. But that perception is out of step with the 70% of Fortune 500 companies that continue to depend on them every single day.
The reality is mainframes power much of our modern world: banks, insurers, governments, retailers, and airlines. In fact, mainframes run 74% of the world’s transactional workloads at just 8% of total IT costs.
So why do finance teams and leadership often question its value? Because a lot of unrelated and adjacent expenses — from shared infrastructure to enterprise software and cross‑platform labor — routinely gets folded into the “mainframe” category. And while it may look like the mainframe is expensive from a sticker-shock perspective, the true cost is often far less than it appears.
In fact, with today’s newer consumption-based pricing models, companies can get cloud-like visibility and control over mainframe costs. But here’s the kicker: despite the rise in adoption of consumption-based pricing, an estimated — and astonishing — two-thirds (65%) of mainframe users still pay fixed, full-capacity rates. A genuine head-scratcher.
Why It Feels Complicated
Mainframe pricing wasn’t always complicated. But over time, systems evolved into the hybrid IT era, and workloads intertwined. What used to be a single system running batch jobs now supports thousands of integrated apps, APIs, and services.
Each of those workloads behaves differently: some spike, others stay steady, and AI or analytics jobs can surge unpredictably. Trying to carve those up under traditional “capacity-based” models get messy fast and forces organizations to attempt unnatural and unnecessary things to manage when and how certain workloads run.
That’s why newer consumption models — like Broadcom’s Mainframe Consumption Licensing and IBM’s Tailor Fit Pricing models — are such a breath of fresh air. They simplify what has become a multilayered pricing puzzle, aligning costs and the way workloads run with actual business value instead of theoretical capacity and pricing constraints.
From Cost to Value
When mainframe teams sit down with finance, the conversation always starts with cost. But it should really start with value.
Under historical models, you paid for total installed capacity — whether it was running production, testing, or sitting idle as backup. You were even charged for disaster recovery sites that weren’t in use. That made compliance-driven requirements like active/active DR setups painfully expensive.
These newer usage-based models flip that. Companies now only pay for what’s actually computed. If production flips from site A to site B, the cost stays the same. If workloads pause, costs pause too. It’s not just fair — it’s smart.
That flexibility has saved organizations millions of dollars over multi-year periods. It’s also given IT leaders a eureka moment: “Wait — we can build resilience and reduce costs?” Yes, you can.
Advice for Untangling Costs
If you’re ready to simplify your mainframe licensing, baseline your current usage. Know what workloads consume the most capacity — and when. Invest in reporting because visibility turns assumptions into facts, especially with sub-capacity reporting tools.
Optimize your workloads. Do more things at once where possible and idle resources that are silent budget drainers. Adopt flexible contracts that let you grow, shrink, or rebalance without penalties. Most importantly, treat the mainframe as part of your hybrid cloud. Not a silo — but a key component of a flexible, modern infrastructure.
Two notable organizations recently achieved this by shifting to consumption-based pricing. The first, a large credit card company, expanded their service internationally without increasing their costs. And a regional bank improved its core apps dramatically, since their vendor provides unlimited usage of all Broadcom mainframe tooling for z/OS application dev and test at no additional cost. In both examples, these winning companies both innovated without the unnatural restrictions associated with traditional mainframe licensing.
Mainframe’s Bottom Line
Yes, mainframe pricing can look intimidating — but it doesn’t have to be. With consumption-based models like Broadcom, IBM, and others, you can finally see what you’re paying for and tie that spend directly to business value.
In other words, mainframe costs don’t need decoding anymore. They just need the right model — one that charges for what you use, rewards efficiency, and keeps your business running at peak resiliency, whether on-prem, in the cloud, or anywhere in between.
Because the future of mainframes isn’t about paying more — it’s about paying smarter and unlocking more value than ever before.