The State of Mainframe Operations.
What 41 operations leaders running production z/OS batch told us about MTTR, on-call load, autonomy adoption and the gap between LLM pilots and real remediation.
The numbers, with the caveats attached.
How the data was collected.
41 operations leaders across BFSI (19), insurance (8), airlines & travel (6), public sector (5), and industrial conglomerates (3). Median fleet: 7 LPARs, 14,200 monthly batch jobs.
Q1 2026 fielding window. Incident metrics aggregated over the trailing 90 days at the time of response, normalized per LPAR.
37-question structured interview plus telemetry export where customers consented (28/41). Telemetry validated against self-reported metrics within ±8%.
Greenfield modernization workloads, IMS-only shops, and air-gapped national-security workloads. These represent material populations and will appear in the 2026 H2 supplement.
Three reading paths.
Where the 73% SME concentration risk shows up first, and which abend families respond to autonomy fastest.
How teams are gating LLM-driven action, the policy-as-code gap, and what the 9% who got it right are doing differently.
The economic model: $14.3K per avoided escalation, sandbox vs enterprise pricing, and a 3-year TCO frame.
Request the full 46-page report.
PDF with sector cuts, all source charts, anonymized interview excerpts, and the response instrument. Distributed under the Progull research NDA.