The warning signs
were already there.

Large programs are often remembered by their final outcome: disruption, delay, abandonment, write-off. The useful information lies earlier—in the signals that were missed, normalized, or separated across functions.

Each analysis preserves the source-grounded record, classifies the outcome, normalizes failure signals and lessons, and compares only what the approved corpus can support.

Editorial independence

Project Failure Intelligence is an independent publication by Steve Thoms, a Microsoft 365 Product Lead with experience in enterprise IT and digital workplace programs. It is not affiliated with the organizations or vendors covered.

Scope

The case base focuses on ERP implementations, CRM programs, infrastructure migrations, and large enterprise transformations. It is designed for pattern recognition and decision support—not blame assignment.