Stage Failure Symptoms

The Five-Stage Cycle has five ways of breaking. Each shows up on the floor before it shows up on a dashboard. This is how to recognize which stage is leaking before the KPI report tells you something has already slipped.

Companion to: Fix the System, Chapter 3. Where to use it: tack this above your desk. Every Monday, read it. Decide which stage is currently leaking worst and spend ten minutes walking the floor against that stage's symptoms.

Stage 01 — IDENTIFY

Symptoms on the floor:

• Operators report issues verbally to a favorite tech ("I'll tell Mike")
• The CMMS backlog looks clean; the hallway backlog is full
• Walk-arounds produce lists that never become notifications
• Rejected notifications disappear without explanation — operators stop writing them

Symptoms in the data:

• Notifications per operator per month is low or trending down
• Triage time from notification creation to first disposition exceeds 48 hours
• The reject pile is not being reviewed monthly

Where the damage shows up downstream: Planning starves. The planner has nothing to plan against but breakdowns and regulatory PMs. Subtle early-warning signals never arrive.

Stage 02 — PLAN

Symptoms on the floor:

• Jobs show up on the floor without a BOM, a procedure, or an hours estimate
• The planner is in breakdown meetings all morning, not at the planning desk
• Technicians discover mid-job that parts are wrong, or the drawing is out of date
• Emergency work gets "planned" in thirty minutes on a Post-it

Symptoms in the data:

• Planned-work ratio below 75%
• Estimated vs. actual hours variance above 40%
• Planner-to-technician ratio is on paper, not in practice

Where the damage shows up downstream: Scheduling becomes a guess. Schedule compliance drops. Operations loses trust. The scheduler stops scheduling against the backlog and starts scheduling around it.

Stage 03 — SCHEDULE

Symptoms on the floor:

• The Monday schedule does not survive Tuesday
• Break-in work runs above 25% of the total
• Operators do not know when the crew is coming because nobody told them
• The weekly scheduling meeting is either not happening, or is ignored by production

Symptoms in the data:

• Schedule compliance below 75%
• Break-in work percent above 20%
• Frozen zone policy exists on paper but not in practice

Where the damage shows up downstream: Execution wrench time craters. Crews context-switch all day. Parts get robbed between jobs (which loops back and breaks planning).

Stage 04 — EXECUTE

Symptoms on the floor:

• Technicians walk to the wrong crib, wait for a shared tool, hunt for a puller
• Crews arrive at an asset before the operator is ready for them
• Vendor meetings and unscheduled tasks hijack leads mid-shift
• Wrench time is below 40% (measure it, don't estimate)

Symptoms in the data:

• Wrench time below Palmer's decent benchmark of 55%
• Parts-related delay percent above 15%
• Shared-tool delay showing up in technician interviews

Where the damage shows up downstream: Close-outs degrade. Technicians have nothing useful to say about jobs that did not go as planned. Scheduling confidence erodes. Planning feedback reflects wait time, not work time.

Stage 05 — CLOSE OUT AND LEARN

Symptoms on the floor:

• Close-out text reads "Repaired. RTO." and nothing else
• Failure codes are not captured, or are captured but never reviewed
• No monthly PM frequency adjustments based on close-out data
• Technicians say "why would I bother writing more? Nobody reads it."

Symptoms in the data:

• Close-out quality audits show <50% of jobs captured failure codes
• PM frequency changes per quarter: zero
• Repeat failures on the same asset with no RCA trigger

Where the damage shows up downstream: The loop is dead. The cycle becomes an open line. Every new turn starts at the operating point of the last — the plant holds or decays, but cannot improve.

Reading the chain backward

A diagnostic trick: the KPI that surfaces the problem is almost never the stage that produces it. Trace the symptom upstream.

The fix is almost always one stage upstream of where the metric lives.

The leakiest stage in most plants

Close-out. Always close-out.

Close-out is the feedback loop that turns the five stages into a cycle. Without it, the cycle is a straight line that cannot learn. Fixing close-out cascades forward into every other stage within a quarter. I have seen plants recover ten points of wrench time in ninety days by doing nothing but rebuilding their close-out rubric and auditing it weekly.

If you are not sure which stage to start with: start with close-out. Restore the loop. Everything else recovers faster with the loop running.

From Fix the System by Ivan Getov. Companion to Chapter 3: The Five-Stage Cycle. getov.xyz/toolkit/fix-the-system/ch03/stage-failure-symptoms. Free. No email required.