Table of contents
Most project “fires” are symptoms of upstream gaps: misaligned decisions, unclear roles, or immature data. Effective root cause analysis is not about blame, but about fixing systems so problems don’t repeat.
Why Symptoms Keep Coming Back
On capital projects, it’s common to treat the visible pain: expedite a PO, add overtime, crash a schedule. These are bandages, not cures. Without addressing the conditions that created the problem, the same issue reappears under a new name a few weeks later.
👉 Effective RCA shifts the team from “Who messed up?” to “What in our system made this outcome likely?”
Signs you’re treating symptoms, not causes:
- The issue “moves” from one workstream to another every month.
- The fix relies on heroics or overtime.
- The corrective action never shows up in your risk register or governance calendar.
A People-First Approach to RCA
Root cause analysis isn’t just a tool; it’s a leadership habit that creates an environment where people can speak openly and succeed without heroics. The posture leaders set determines whether RCA drives learning or blame:
- Curiosity over blame. Engineers and planners must feel safe to speak to facts without fear of punishment..
- Evidence over opinion. Logs, dates, and approvals should be visible for all.
- Systems over superheroes. If success depends on one person working miracles, the system hasn’t been fixed and people are being set up to fail..
With that foundation, use the The Three Levels test to target the right level of failure.
The Three Levels of Failure
Most repeat problems trace back to policy or supervisory gaps, not individuals.
- Policy / Corporate Level: Is there a clear, current, workable policy? Is it aligned with incentives? If not, the failure is systemic.
- Supervisory Level: Were people trained, resourced, and supported to follow the policy? Was it enforced consistently?
- Individual Level: Did a trained, supported person willfully disregard? True individual root causes are rare.
👉 Rule of Thumb: If your RCA pins the cause on one person, you probably haven’t found the root cause yet.
Symptom vs. Root Cause: Quick Contrasts
| Symptom (what you see) | Likely Root Cause | Better Fix |
|---|---|---|
| Late vendor datasheets | Scope churn after FEL-2; unclear approval authority | Freeze scope, clarify RACI, decision log |
| Field rework on piping | Incomplete P&IDs at authorization; late model reviews | Definition maturity criteria before gate; model review cadence |
| Procurement congestion | Bulk quantities unstable; late change propagation | Quantity control plan; formal change control |
Proof Point: The Vendor Datasheet Problem
Context: A $250M brownfield expansion faced chronic vendor submittal delays threatening steel fabrication.
- Symptom: Datasheets consistently 2–3 weeks late.
- RCA Path: Vendor was waiting on final motor specs; package engineer unclear on approval authority; two parallel approval paths (project vs. plant) with no single accountable.
- Root Cause: No single-point accountability for spec approvals; a governance gap from FEL-2.
- Fix: One approval authority named, RACI updated, weekly “spec board” showing queue and aging.
- Result: Datasheet lateness dropped from 14 days to 3 days in one month.
What Good Looks Like in Practice
- Right-sized: Quick huddles for recurring issues; deep dives only for safety or cross-system failures.
- Systemic fixes: Countermeasures change workflows, policies, or incentives, not just individuals.
- Visible in governance: Actions logged in the risk register, tracked in gate packs, and measured in KPIs.
- Cultural shift: Fewer heroics, fewer repeats, steadier delivery.
Closing Thought
Most project “fires” aren’t fixed with heroics or overtime, they’re prevented by addressing the system gaps that made them inevitable. Root cause analysis, when done well, ties each failure back to definition maturity, governance, or supervisory discipline and embeds the fix into normal project controls. That’s what stops the same issue from resurfacing under a new name. Projects rarely fail because someone slipped once; they fail because leaders let the same conditions repeat. Fix the system, and the outcomes will follow.
