Manufacturing Dashboard KPIs That Matter (And 7 Vanity Metrics That Don’t)

Manufacturing dashboard KPI selection is one of the least rigorous activities in most plants. The typical process: a committee lists every metric someone has ever tracked, proposes adding them to a dashboard, and the dashboard becomes a 20-KPI wall of numbers that nobody actually reads. The result is what dashboard designers call vanity metrics — numbers that feel important, look professional, but drive no decisions. Vanity metrics dilute the actionable KPIs around them because screen real estate is finite and attention is finite. A dashboard with 5 actionable KPIs plus 15 vanity metrics functions worse than a dashboard with only the 5 actionable ones, because the vanity metrics obscure the signal.

This article catalogs the manufacturing KPIs that consistently drive decisions across 450+ deployments and the vanity metrics that consistently do not. The distinction is not about data accuracy or technical sophistication; all the KPIs and vanity metrics discussed can be measured accurately. The distinction is whether the number, when it crosses a threshold, triggers a specific action by a specific person. KPIs that do: keep. Metrics that do not: remove from the dashboard, even if they are measured and reported elsewhere.

The 7 manufacturing KPIs that matter

1. OEE (Overall Equipment Effectiveness). The single most actionable manufacturing KPI when measured honestly. Triggers operator escalation, supervisor coaching, engineering investigation, capital investment decisions. Every operational audience has a decision that maps to OEE at specific thresholds. The key caveat is measurement honesty: inflated OEE (10-18 points above reality) produces false reassurance rather than action.

2. Top Pareto cause (current shift). A text KPI naming the #1 loss category for the current shift, with count. Operators use it to target immediate attention; supervisors use it to identify recurring issues across shifts; engineering uses it to prioritize root-cause investigations. The Pareto KPI is what makes OEE actionable rather than merely measurable.

3. First-pass yield (quality rate without rework). Unambiguous quality signal that drives immediate action when below threshold. Different from “shipped quality” (which includes reworked parts) because it captures true process capability. Operators use it to pause lines when quality drops; supervisors use it to trigger quality holds and investigations.

4. Changeover time (current run). Measured during changeovers with live countdown. Drives operator focus during changeover execution; drives SMED program focus for supervisors; drives capital planning for executives. A single KPI that meaningfully serves all three audiences when scoped correctly.

5. Stop count (rolling 2-hour window). Counts stops in a rolling 2-hour window, distinct from total shift stops. The rolling window makes it actionable: an unusual spike is visible immediately instead of buried in shift totals. Drives operator engineering-intervention decisions and supervisor coaching.

6. Schedule adherence (planned vs actual completion). Primary KPI for production planning. Drives operational decisions about prioritization, overtime authorization, customer commitment escalation. Usually a supervisor and planning-team KPI rather than operator, but critical for those audiences.

7. MTBF and MTTR (reliability metrics). Mean Time Between Failures and Mean Time To Repair. Engineering and maintenance KPIs that drive preventive maintenance scheduling, spare parts stocking, reliability investment decisions. Less relevant for operators but critical for the maintenance engineering audience.

The 7 vanity metrics that don’t

Vanity 1: Total production output (cumulative count). Increases monotonically through the shift. The final number is useful for reporting; the running cumulative count during the shift drives no decisions. Replace with “throughput vs target” which shows delta from plan in real time.

Vanity 2: Running percentage (time machine was running vs total calendar time). Not a loss-category signal. A 67% running percentage tells you little because it combines planned-downtime time with unplanned-downtime time into one number. Replace with Availability (within OEE) which isolates the unplanned.

Vanity 3: Total number of parts produced this month. Monthly cumulative metrics are not actionable at the operator or supervisor level. They belong in monthly reports, not real-time dashboards. Replace with current-shift throughput vs target.

Vanity 4: Average OEE for the month. Monthly averages obscure recent performance because they include distant data points. Weighted more toward recent weeks is marginally better but still obscures current-shift trends. Replace with OEE trend chart showing last 30 days for supervisor context.

Vanity 5: Number of products in portfolio. Some dashboards display the number of SKUs the plant produces. It does not drive decisions; it is a plant attribute, not a KPI. Remove entirely.

Vanity 6: Year-over-year comparison for operational metrics. YoY comparison is appropriate for executive strategic metrics but distracts from operational dashboards. An operator does not benefit from knowing last year’s shift performance; they need this-shift targets and current trends.

Vanity 7: Equipment utilization displayed as a percentage. Without context, utilization percentages mean nothing. Is 72% utilization good? Depends on the plant’s operational model (2 shifts vs 3 shifts, continuous vs intermittent production). Replace with specific actionable metrics: throughput vs target, availability, OEE.

Industry-specific KPI additions

Beyond the core 7 KPIs that apply across discrete manufacturing, specific industries require additional KPIs tied to their regulatory or operational context. Pharmaceutical packaging: serialization exception count, batch record completion, inline vision reject rate. These replace or augment quality-rate metrics because regulatory context makes them consequential. Automotive tier-1: first-time capability index (Cpk) on critical features, customer-specific containment activations, PPAP completion status. Customer quality requirements drive specific metric needs. Food and beverage: CIP completion status, allergen changeover verification, temperature excursion count. Food safety regulations drive measurement requirements. Chemicals and specialty: energy consumption per batch, solvent recovery rate, ventilation differential pressure. Process-specific metrics that dominate operational decisions.

The industry-specific additions should follow the same selection discipline as the core 7: each must map to a specific decision by a specific audience at a specific threshold. Industry metrics added to dashboards without decision mapping become industry-flavored vanity metrics.

How many KPIs total on a dashboard

Operator dashboards: 5 KPIs (the 5-KPI rule). Supervisor dashboards: 10-15 KPIs across multiple lines or dimensions. Executive dashboards: 6-8 KPIs across plants or business units. Going above these ranges produces cognitive overload and reduces dashboard effectiveness. Going significantly below misses metrics the audience genuinely needs.

The discipline is hard for committees of stakeholders, each of whom wants their preferred metric displayed. The resolution is to give each stakeholder the detailed report they need separately from the dashboard, not to pile their metrics onto the dashboard. Dashboards are for ambient decision-driving at the right frequency; reports are for periodic deep review. Confusing the two functions corrupts both.

The 30-day KPI validation process

Before committing to a dashboard design, run a 30-day KPI validation on a pilot audience. For each proposed KPI, log every instance where a user looks at the dashboard, which KPI they focus on, what decision results from viewing it. At 30 days, tally which KPIs drove decisions and which did not. Remove any KPI that had no decision in 30 days. Add any KPI the audience requested mid-pilot because they lacked it.

The validated dashboard after 30 days is typically 40-60% smaller than the initial design, with the remaining KPIs carrying dramatically more visual weight. Utilization rates post-validation typically improve 2-3x versus the pre-validation design.

Start with actionable dashboards from day one — Free 48-hour POC
Validated KPI selection, operator/supervisor views, direct-sensor data, decision-ready from deployment
Request a free TeepTrak POC

External references: Wikipedia: KPI · MESA International · NIST

See also: Manufacturing Dashboard Design Guide 2026 · Real-Time vs Daily: Refresh Rate Decision · OEE Software Overview