Sunday, June 21, 2026

IT-org redesign — defining the "strategic PMO" operating model in the AI era

Sara Gallagher argues CIOs calling for "strategic PMOs" must first answer six operating-model questions — purpose, structure, people, process, tools, and culture — since AI agents are now absorbing coordination, analysis, and reporting work traditionally done by PMs, and a PMO that merely tracks status doesn't protect business cases the way one that flags changed assumptions and surfaces portfolio tradeoffs does.

The piece flags two structural gaps: a governance gap (how humans provide meaningful oversight when AI agents scope, prioritize, and route work faster than review capacity allows) and a data-readiness gap (most organizations' project artifacts — scattered files, Word docs, bloated meeting transcripts — aren't AI-ready), plus a shift toward hybrid SaaS pricing bundling per-seat human access with metered agent-consumption credits.

Watch whether CIOs move past "AI literacy" platitudes toward concrete day-in-the-life PM workflow redesigns, and watch vendor enterprise agreements (ServiceNow, Atlassian, Microsoft) continue restructuring around agent-credit consumption rather than flat per-seat pricing.

CFO/P&L-level AI-ROI measurement and budget governance

A Bain survey of 951 global companies found AI cost savings are consistently falling short of targets — for example, 37% of companies targeted 11-20% savings but only 29% achieved that range — yet 90% of the same companies are increasing AI budgets again, now directed at more autonomous, complex AI agents rather than the simpler use cases that underdelivered.

The leading barrier to AI progress, cited by 41% of respondents, was inadequate data access and integration across systems — not model capability — underscoring that the savings shortfall is largely an enterprise-data and integration problem rather than an AI-quality problem.

Watch whether the same data-integration barrier reappears as the binding constraint on ROI from the next wave of more autonomous agent deployments, and watch if CFOs start tying AI budget increases to integration-readiness milestones rather than tool adoption alone.