MM
Mareeswar Muthiah
Mar 2026 · 6 min read · Modernisation

The TCO model has shifted

For two decades, mainframe modernisation was an exercise in faith. The numbers were always close to neutral on a five-year horizon, and CIOs would (rationally) defer the spend. That shifted in 2024 — and the gap is now wide enough that the financial case writes itself.

What changed

Three things changed at once: cloud-native databases finally caught up to MIPS-grade workloads, the talent market for mainframe operators tightened to the point where retention is unaffordable, and CDC tooling (Debezium, Kafka Connect) matured to a level where dual-write is no longer a research project.

The 22-month payback

On the seven core modernisation programmes we have shipped since 2023, the median payback period was 22 months. That includes the migration cost, the parallel-run cost, and a 15% buffer for in-flight risk. In the most favourable case it was 14 months; in the toughest, 31.

How to design the business case

Anchor the case on three lines: (1) run-cost compression year over year, (2) time-to-market for new products, and (3) talent-cost avoidance for the next decade. We have rarely seen a modernisation business case fail at the second item.

Where to start

The strangler-fig pattern. Pick the smallest domain that produces externally visible value, build it side-by-side, run it live-parallel for a meaningful window, and only then cut over. Treat any “big bang” cutover proposal as a red flag in the steering committee.

Talk to a partner

Want to discuss this?

A senior partner will respond within one business day.