WHY MOST PLANTS TRACK TOO MANY KPIS
The standard path to a KPI problem in manufacturing is well worn. Leadership decides metrics matter. A committee assembles. Every function advocates for the metrics it cares about. Finance wants cost variance and gross margin. Quality wants first pass yield, PPM, customer complaints, and supplier reject rate. Maintenance wants MTBF, MTTR, PM compliance, and OEE. Safety wants recordable rate, days since last incident, near-miss count, and safety observation completion. HR wants turnover, attendance, and open positions.
Three months later, the plant is tracking 45 metrics on a dashboard that nobody has time to review completely. Metrics get updated sporadically. Leadership looks at the ones they already understand and ignores the rest. The metrics that would require action go unreviewed, and the action does not happen.
The problem is not that any individual metric is wrong. Most of them are reasonable things to know. The problem is that a metric nobody acts on is not a management tool. It is administrative overhead. Every metric on a dashboard has a cost: someone's time to collect it, someone's time to review it, and the opportunity cost of the attention that the relevant but unreviewed metric never received.
THE TRAP OF MEASURING WITHOUT ACTING
A plant that tracks scrap rate monthly but has not launched a scrap reduction project in 18 months is not managing scrap. It is measuring it. The difference is not a question of data quality or dashboard design. It is a question of whether the metric connects to a decision and an owner.
Every metric worth tracking needs two things: an owner who is accountable for the result, and a decision that changes based on how the metric moves. If a metric can be below target for three consecutive months without triggering a specific action by a specific person, it should not be on the dashboard.
This is not a reason to stop measuring. It is a reason to be disciplined about which metrics make the cut. A dashboard with 10 metrics that every leader understands, reviews at the right frequency, and acts on when they move is more valuable than a dashboard with 50 metrics that produces monthly reports nobody reads. The three numbers by 10am post covers the daily version of this principle: what is the minimum set of metrics that tells you whether today is on track?
SAFETY KPIS: THE LAGGING AND LEADING PAIR
Recordable incident rate (RIR). OSHA recordable incidents per 100 full-time employees, calculated over a rolling 12 months. This is the lagging safety indicator. It tells you where you have been. Target varies by industry; any number above zero warrants investigation of the contributing factors.
Near-miss count. The number of near-miss events reported in the trailing month. This is the leading safety indicator. A high near-miss count with a low RIR is a healthy safety culture: people are seeing and reporting hazards before injuries happen. A low near-miss count in any plant with significant floor activity means the near-miss system is not working, not that the plant is safe. Near-miss count should trend up in a plant that is actively building its safety culture.
QUALITY KPIS: INTERNAL PERFORMANCE AND EXTERNAL SIGNAL
First pass yield. The percentage of parts or units that pass all quality checks on the first attempt, without rework or rejection. Measured at the part or product family level, and ideally at the operation level where the loss originates. A plant with 94 percent first pass yield is spending resources on 6 percent of its production twice. The manufacturing scrap tracking post covers how to build the operation-level data that makes first pass yield actionable rather than just reportable.
Customer escape rate. The number of customer quality complaints or returns per period, normalized by shipped volume. Internal quality metrics (first pass yield, scrap rate) tell you about your internal process. Customer escapes tell you what got through. Both matter, and a gap between the two (good internal metrics, poor customer escapes) is a signal that your inspection is not catching what your customer is finding.
DELIVERY KPIS: INTERNAL RELIABILITY AND CUSTOMER RESULT
Schedule attainment. The percentage of jobs or orders completed on the planned date within the scheduling period. Measured weekly. A plant hitting 85 percent schedule attainment is missing nearly one in six jobs from the plan. Schedule attainment is the leading indicator of on-time delivery: if jobs are completing on schedule internally, they ship on time. If they are not, expediting and premium freight fill the gap. The manufacturing production scheduling post covers how schedule attainment is built or destroyed in the scheduling process.
On-time delivery (OTD). The percentage of customer orders shipped on or before the customer-required date. This is the external delivery metric. Schedule attainment is the internal early warning. OTD is the customer-facing result. A plant that chronically expedites can maintain high OTD while schedule attainment is poor, which means overtime and expediting costs are hiding the scheduling problem rather than fixing it.
COST KPIS: PRODUCTIVITY AND PREMIUM TIME
Cost per unit. Total manufacturing cost divided by units produced in the period. When cost per unit rises, something in the manufacturing process is less efficient than the plan expected: labor productivity is down, scrap is up, overhead is being spread over fewer units, or purchased material costs increased. Tracking it weekly identifies the change faster than monthly P&L review. The manufacturing P&L post covers how the operational metrics that drive cost per unit connect to the income statement lines where they show up.
Overtime percentage. Overtime hours as a percentage of total hours worked, by shift and by week. The trend matters as much as the absolute number. A plant where overtime consistently runs above 10 percent has a structural scheduling or staffing problem. A sudden spike in overtime is a signal of a specific event (equipment failure, quality miss, demand surge) that warrants investigation.
PEOPLE KPIS: THE WORKFORCE STABILITY SIGNAL
Voluntary turnover rate. The number of voluntary separations in the trailing 12 months divided by average headcount, expressed as a percent. In a plant with 60 employees, a 30 percent annual turnover rate means replacing 18 people per year. The fully loaded cost of that turnover typically runs $3,000 to $8,000 per replacement. Track it quarterly and trend it over 12 months. Sudden increases often correlate with management changes or compensation market moves.
Attendance rate. The percentage of scheduled shifts worked across all direct labor employees in the trailing month. Absenteeism above 3 to 4 percent starts to affect scheduling reliability and forces overtime to cover open spots. Tracking attendance at the shift level reveals patterns that an aggregate number hides: a third-shift attendance problem looks fine in the monthly aggregate and is disruptive in the daily schedule.
WHAT NOT TO MEASURE AT THIS SIZE
Several metrics appear frequently on manufacturing dashboards and belong off of them for most small and midmarket plants.
Theoretical cycle time versus actual is a useful engineering tool for bottleneck analysis. It is not a useful management metric because it compares actual performance to an unachievable ideal rather than to a meaningful operational target.
Productivity index without current standards. A productivity index (output per labor hour relative to a standard) is only meaningful if the standard is current. Most plants whose standards have not been updated in two or more years are measuring variance from an obsolete baseline, which produces numbers that are not interpretable.
More than two financial metrics at the supervisor or operator level. Cost per unit and overtime percentage are sufficient at this level. Gross margin, SG&A variance, and overhead absorption belong in the monthly business review. They are not useful on a daily production board.
HOW CADENCE DETERMINES WHICH KPIS GO WHERE
The right metric at the wrong frequency is noise. Safety near-miss count reviewed annually tells you nothing actionable. On-time delivery reviewed hourly is too granular for the signal it carries. Match the metric to the decision frequency.
Daily: output versus plan, safety observations or near-miss events, any quality hold events. Reviewed at the start-of-shift meeting or the daily production meeting.
Weekly: schedule attainment, first pass yield by operation, overtime percent, near-miss count. Reviewed in the weekly operations meeting.
Monthly: recordable rate, on-time delivery, customer escapes, cost per unit, voluntary turnover, attendance rate. Reviewed in the monthly business review.
Every metric in the essential 10 fits into one of these cadences. The rule: every KPI needs an owner who is present at the review where that KPI is discussed, and a defined response when the metric misses target.
P4 Daily Management is one of the ten pillars in the Sharpen 10-pillar framework. A plant where the right metrics are reviewed at the right frequency by the right owners is the operational definition of a functioning management system. Without that structure, even well-chosen KPIs become decorative.
WHAT TO DO NEXT
Start with the 10 essential KPIs. For each one, confirm the data source, assign an owner, set a target, and define what action triggers when the metric misses. Build a review cadence that covers all 10 within the monthly cycle. Add nothing to the dashboard until the core 10 are running cleanly with consistent owners and consistent follow-through.
The free Sharpen diagnostic at /intake takes about 10 minutes and produces a prioritized roadmap across all ten operational pillars. If daily management discipline or financial visibility is a gap, the diagnostic will surface it and help sequence the work.