IBM Mainframe Licensing Review and Optimization for a Financial Institution

The Challenge

The institution in question is a mid-tier European regional bank with approximately 6,000 employees across eight countries, including operations in Germany, France, Netherlands, Belgium, Austria, Switzerland, Sweden, and the United Kingdom. Its core banking infrastructure runs on IBM z/OS mainframe systems, which process mission-critical transaction volumes, settlement operations, and regulatory compliance workflows. The bank had faced escalating IBM software costs year-on-year, driven partly by genuine transaction volume growth but significantly exacerbated by suboptimal licensing structure decisions made four years prior.

When Redress conducted the initial assessment, the bank's total annual IBM mainframe software spend reached approximately €22 million. The primary drivers of cost escalation were three-fold: first, the bank's Measured Usage Licence (MUL) pricing structure had not been recalibrated since inception, despite material changes in transaction patterns and workload distribution; second, zIIP (IBM System z Integrated Information Processor) specialty engines—designed specifically to execute eligible database and analytics workloads at dramatically reduced license cost—were significantly underdeployed, running at only 14% of eligible offload capacity; and third, Software Subscription and Support (S&S) fees had compounded annually for four years without formal renegotiation, despite the bank's substantial scale and vendor switching costs.

The institution's IT leadership recognized the inefficiency but lacked the technical depth and commercial leverage to challenge IBM's existing commercial terms. Additionally, regulatory modernization pressures—including PSD2 compliance, GDPR data governance, and capital adequacy reporting—had forced prioritization of feature enhancements and regulatory systems over cost optimization, leaving the licensing function underfunded and understaffed.

The Approach

Redress deployed a three-phase engagement model: diagnostic assessment, optimization strategy, and renegotiation execution.

Phase 1: Diagnostic Assessment

The team conducted a 12-week technical and commercial audit covering: (1) z/OS LPAR configuration review and historical capacity growth analysis; (2) detailed MIPS (Million Instructions Per Second) measurement against actual monthly transaction volumes for the past 36 months, identifying significant over-provisioning during non-peak periods; (3) workload analysis to identify eligible offload candidates currently executing on general-purpose processors; (4) comparative benchmarking against peer financial institutions of similar scale and geographic footprint; and (5) comprehensive Software Subscription & Support invoice analysis spanning four years, identifying inconsistencies and duplicate charges.

Key diagnostic findings included: the bank's provisioned MIPS was calibrated to historical peak capacity rather than 99th percentile actual demand, creating 22% structural over-capacity; approximately 38% of database and transaction processing workloads were eligible for zIIP offload but executing on general-purpose engines due to legacy application configuration and insufficient zIIP capacity; and S&S pricing had been applied retroactively to legacy license acquisitions, compounding annual costs by €280K annually.

Phase 2: Optimization Strategy

The optimization strategy comprised three concurrent workstreams:

Workstream A: MUL and MIPS Right-Sizing (€2.9M Annual Savings)

Redress recommended a 15% reduction in provisioned MIPS capacity, from 9,200 MIPS to 7,850 MIPS, based on three-year actual peak utilization metrics. This was structured as a rolling contract amendment rather than immediate license reduction, allowing the bank to manage capacity risk while capturing savings. Additionally, the MUL structure was renegotiated to incorporate variable tier pricing aligned to actual monthly MIPS consumption, reducing fixed-cost exposure and incentivizing further workload optimization. The net effect was €2.9M in annual savings, with the first-year benefit realized within 60 days of contract execution.

Workstream B: zIIP Specialty Engine Expansion (€1.8M Annual Savings)

The bank's existing zIIP configuration supported only 1,200 MIPS of offload capacity. By re-architecting database workload distribution and upgrading the zIIP processor configuration to 3,100 MIPS capacity—at a capital cost of €650K amortized over three years—Redress enabled the bank to expand zIIP offload from 14% to 52% of eligible workload, equivalent to 2,760 MIPS of traditional processor capacity retired. Given IBM's aggressive zIIP pricing advantage (typically 12-18% of general-purpose license cost), this expansion generated €1.8M in annual software cost avoidance and improved system performance due to reduced general-purpose processor contention.

Workstream C: Software Subscription & Support Renegotiation (€1.1M Annual Savings)

The bank's S&S obligations had compounded due to outdated contract language and lack of vendor governance. Redress identified an aggregate €1.1M in annual overcharges through: (1) removal of S&S obligations on permanently retired legacy software (€380K); (2) renegotiation of support tier levels aligned to actual service consumption metrics (€420K); and (3) consolidation of redundant support contracts across eight geographic operations (€300K). The renegotiation was executed as a six-month amendment cycle, allowing the bank to phase in governance improvements without disruption.

The Outcome

Total realized savings across all three workstreams exceeded €5.8M annually, representing a 26% reduction in baseline mainframe software spend. These savings were achieved without service disruption, regulatory compliance compromise, or material capital investment beyond the zIIP upgrade (which itself generated a 2.7-year payback period).

The bank implemented a formal mainframe licensing governance framework post-engagement, establishing quarterly cost reviews, automated MIPS consumption reporting, and a workload optimization steering committee. This governance structure has enabled an additional €800K in incremental savings through continuous optimization in the 12 months following the primary engagement, without further vendor renegotiation.

By quarter 3 of implementation, the bank had realized 92% of projected savings, with full run-rate achieved by quarter 4. The engagement also provided a blueprint for cost optimization across the institution's broader enterprise software portfolio (Oracle, SAP, Broadcom), ultimately influencing multi-year technology investment strategy and vendor relationship governance.

Key Takeaways

• Measured Usage Licenses require regular recalibration: MUL structures are not static. Actual workload patterns, capacity utilization, and transaction volumes evolve materially over time. Annual or biennial MUL benchmarking—benchmarked against peer institutions and industry demand patterns—is essential to avoid structural over-provisioning and the compounding cost impact across multiple-year contracts.
• Specialty processor offload is often vastly underdeployed: zIIP, Integrated Facility for Linux (IFL), and other specialty processors are frequently enabled but underutilized due to legacy application design or insufficient processor capacity planning. A systematic workload audit can identify 25-40% of eligible offload candidates in mature mainframe environments, with implementation ROI typically under three years.
• Software Subscription & Support governance is often absent: S&S charges are frequently opaque and subject to vendor interpretation. Establishing formal governance—including invoice validation, retirement of unused software, and periodic rate benchmarking—typically identifies 8-12% cost reductions without contract renegotiation.
• Financial institutions should formalize mainframe cost ownership: Given the criticality and complexity of mainframe systems, cost governance should be a shared responsibility across IT operations, architecture, finance, and vendor management. Quarterly steering committees with clear accountability improve both cost outcomes and strategic alignment.
• Vendor engagement should emphasize mutual value creation: The most successful optimizations position the bank's cost reduction as aligned with IBM's strategic interests (e.g., higher z/OS utilization, better zIIP penetration, reduced support overhead). This framing enables genuine commercial collaboration rather than adversarial negotiation.