In March 2024, a suspected arson attack at Tesla’s Grünheide plant near Berlin knocked out power for an entire week. Production halted. Deliveries stalled. The cost: more than €100 million in seven days (ISM, 2024).
Tesla CEO Elon Musk had already described idle factories as “money furnaces” during a 2022 earnings call, and the Berlin incident proved why: when production stops, the financial bleed is immediate, visible, and devastating.
That single event illustrates what Siemens quantified across the entire Fortune Global 500: unplanned downtime now drains 11% of annual revenue from the world’s largest manufacturers—$1.4 trillion combined (Siemens True Cost of Downtime, 2024).
The only defense against losses at this scale is a rigorous system of key production indicators that detect deterioration before it becomes a shutdown.
| What to Remember |
| Unplanned downtime now costs Fortune Global 500 manufacturers 11% of annual revenue—a combined $1.4 trillion (Siemens, 2024). The hourly cost ranges from $39,000 in consumer goods to $2.3 million in automotive. Key production indicators are the early-warning system that prevents these losses. |
| Only ~3% of manufacturing plants globally achieve world-class OEE (85%+). The global average sits between 55–60% (Evocon, 2024). This gap represents massive improvement opportunity—a plant moving from 60% to 75% OEE can unlock 20–40% more throughput without capital investment. |
| Unplanned downtime accounts for 34.2% of all efficiency losses in discrete manufacturing, followed by setup/changeover time at 28.7% and material shortages at 18.4% (Godlan, 2025). Tracking these loss categories through structured KPIs is the first step to eliminating them. |
| The most effective KPI programs limit themselves to 8–12 core metrics, organized by operational (OEE, cycle time, throughput), financial (cost per unit, scrap rate), and customer-facing (on-time delivery, first pass yield) categories. More than 12 KPIs dilutes focus and overwhelms frontline teams. |
| A risk-based approach to production KPIs—mapping each metric to potential failure modes using ISO 31000 principles—transforms reactive firefighting into proactive risk management. Manufacturers who treat KPI programs as risk controls outperform those who treat them as reporting exercises. |
This guide builds that system—grounded in benchmark data from 1,470+ manufacturing operations, mapped to a risk management framework that treats every KPI as a risk control.
Figure 1: OEE benchmarks vary widely by industry, with only ~3% of global plants achieving world-class status (85%+).
Key Production Indicators vs. General KPIs: Drawing the Right Boundary
Every metric in manufacturing is not a key production indicator. The distinction matters because tracking 50 metrics creates data noise; tracking 10 well-chosen key risk indicators creates actionable intelligence.
Key production indicators are the subset of manufacturing KPIs that directly measure the efficiency, quality, and throughput of the production process itself—not sales, not marketing, not HR.
They answer three questions: Are we producing at the right speed? Are we producing the right quality? Are we using our assets effectively?
General KPIs like revenue growth, customer acquisition cost, or employee satisfaction serve important strategic purposes, but they sit at a different level of the business hierarchy.
Production indicators feed into those outcomes. When your OEE drops from 72% to 58%, it will eventually show up as missed deliveries, higher unit costs, and declining customer satisfaction—but by the time those lagging indicators move, weeks of production value have already been lost.
The lesson from leading vs. lagging KRIs applies directly: production indicators are leading signals. Strategic KPIs are lagging confirmations.
The Three Tiers of Manufacturing Measurement
| Tier | Metrics | Decision Level |
| Operational (Shop Floor) | OEE, cycle time, throughput, scrap rate, changeover time, machine availability | Shift supervisors and production managers. Real-time decisions about line speed, maintenance, and staffing. |
| Tactical (Plant Level) | On-time delivery (OTIF), cost per unit, WIP levels, first pass yield, planned vs. unplanned maintenance ratio | Plant directors and operations VPs. Weekly/monthly decisions about resource allocation, process improvement, and capacity planning. |
| Strategic (Enterprise) | Revenue per employee, total manufacturing cost as % of revenue, capacity utilization trend, customer complaint rate | C-suite and board. Quarterly/annual decisions about capital investment, plant expansion, and market strategy. |
The Five Production Indicators Every Manufacturer Must Track
With the tier framework established, we can now identify the specific key production indicators that deliver the highest signal-to-noise ratio.
These five metrics, when tracked consistently, provide a near-complete picture of production health. Each maps directly to a specific risk assessment domain. For process-specific examples, see our injection moulding risk assessment guide.
1. Overall Equipment Effectiveness (OEE)
OEE combines three sub-metrics—Availability, Performance, and Quality—into a single percentage that reveals how much of your planned production time is truly productive.
A plant running at 90% availability, 90% performance, and 90% quality achieves only 72.9% OEE, not 90%. The multiplicative effect is what makes OEE so powerful: it exposes hidden losses that individual metrics miss.
Godlan’s 2025 benchmark study of 1,470+ discrete manufacturing operations found that medical devices led at 78.2% average OEE, driven by regulatory compliance requirements that enforce process discipline.
Typical discrete manufacturers average 60%, and Evocon’s global dataset (3,500+ machines across 50+ countries) confirmed that only approximately 3% of plants consistently achieve world-class OEE of 85%+ (Evocon, 2024).
Pharmaceutical manufacturers average just 35% OEE, largely due to cleaning validation and batch changeover requirements (SCW.ai, 2025).
The practical implication: a pharmaceutical plant that moves from 35% to 60% OEE could unlock 20–60% more throughput with zero capital investment—translating to an estimated $14–$16 million annual return for a 50-line facility (SCW.ai calculation).
That makes OEE improvement one of the highest-ROI initiatives in manufacturing.
Figure 2: Unplanned downtime and setup time together account for nearly two-thirds of all efficiency losses.
2. Cycle Time
Cycle time measures the elapsed time from the start of a production operation to its completion. Shorter cycle times mean higher throughput capacity, but only if quality holds. Tracking cycle time alongside first pass yield prevents the common trap of speeding up production at the expense of scrap rates.
The relationship between these two metrics is a textbook risk treatment tradeoff: optimizing one without monitoring the other creates a new risk.
3. Throughput
Throughput measures units produced per unit of time. Unlike cycle time (which tracks a single operation), throughput captures the end-to-end production rate including all wait times, bottlenecks, and transfer delays between stations.
A plant with excellent individual cycle times but poor throughput has a constraint somewhere in its value stream—and Eliyahu Goldratt’s Theory of Constraints provides the framework for finding it.
4. First Pass Yield (FPY)
First pass yield measures the percentage of units that pass quality inspection on the first attempt, without rework or rejection.
An FPY of 95% sounds strong until you recognize that across a 10-step production process, the rolled throughput yield (FPY at each step multiplied together) drops to 0.95^10 = 59.9%. Tracking FPY at the process level, not just the final inspection, reveals quality erosion that aggregate metrics conceal.
5. On-Time-In-Full (OTIF) Delivery
OTIF measures the percentage of customer orders delivered complete and on schedule. This is the customer-facing indicator that connects production performance to commercial outcomes. World-class OTIF sits above 95%, but many manufacturers operate in the 80–90% range.
Every percentage point of OTIF improvement reduces expediting costs, strengthens customer retention, and shortens the cash cycle.
From a business continuity perspective, OTIF is also a critical dependency: when OTIF drops below contractual thresholds, supply chain partners may activate penalty clauses or seek alternative suppliers.
Complete Key Production Indicators Reference Table
The table below compiles the essential manufacturing KPIs with their formulas, benchmark targets, and the specific risk each metric monitors. Use this as a risk register for your production KPI program.
| KPI | Formula | World-Class Target | Typical Average | Risk Monitored |
| OEE | Availability × Performance × Quality | 85%+ | 55–60% | Asset utilization loss |
| Cycle Time | End time – Start time per unit | Varies by product | Varies | Throughput bottleneck |
| Throughput | Total units / Time period | At or above planned rate | 80–90% of plan | Capacity underutilization |
| First Pass Yield | (Good units at first pass / Total units) × 100 | 95%+ | 85–90% | Quality cost and rework |
| OTIF Delivery | (Orders delivered on-time and in-full / Total orders) × 100 | 95%+ | 80–90% | Customer attrition |
| Scrap Rate | (Scrapped units / Total units) × 100 | <2% | 3–5% | Material waste cost |
| Changeover Time | Time from last good unit (Product A) to first good unit (Product B) | Single-digit minutes (SMED) | 30–90 min | Flexibility and throughput |
| Planned Maintenance % | (Planned maintenance hours / Total maintenance hours) × 100 | 85%+ | 50–70% | Unplanned downtime risk |
| WIP Levels | Count of units in process at any point | Minimal (pull system) | Often excessive | Cash flow and lead time |
| Cost Per Unit | Total production cost / Total units | Below industry benchmark | Varies by sector | Profitability erosion |
Figure 3: Automotive downtime costs have risen 50%+ since 2019, now exceeding $2 million per hour.
Quantifying the Stakes: What Production KPIs Are Protecting
Abstract discussions about “efficiency” and “productivity” lose urgency without dollar figures attached. The Siemens True Cost of Downtime reports (2022 and 2024) provide the most comprehensive financial analysis of manufacturing downtime ever published, and the numbers should focus every operations leader’s attention:
The average cost of one hour of unplanned downtime in manufacturing is $260,000 (Aberdeen/ServiceMax). In automotive, that figure reaches $2.3 million per hour—up over 50% from $1.3 million in 2019–20. Oil and gas downtime costs have more than doubled to nearly $500,000 per hour.
The average manufacturer faces approximately 800 hours of unplanned downtime annually, equivalent to about 15 hours per week of paid non-productive time (Siemens, 2024).
Equipment failures account for roughly 42% of that downtime. The remaining 58% comes from material shortages, changeovers, quality holds, and process issues—every one of which is detectable through the key production indicators outlined above.
Manufacturers who track OEE components, planned maintenance ratios, and WIP levels have the early-warning system needed to address these losses before they cascade into unplanned stoppages.
Figure 4: Downtime costs are rising faster than inflation across all manufacturing sectors.
Applying Risk Management Frameworks to Production KPIs
Most manufacturing KPI programs are built as performance dashboards. That’s necessary but insufficient. The practitioners who extract maximum value from production indicators treat them as risk controls within a structured enterprise risk management framework. The difference is not semantic; it changes how you set thresholds, who owns escalation, and what happens when a metric breaches its limit.
The KRI Approach to Production Metrics
Reframing KPIs as key risk indicators (KRIs) means assigning each metric a three-tier threshold structure: green (within tolerance), amber (approaching limit—investigate), and red (breach—escalate immediately). This is the same risk appetite structure used in financial services, and it works just as effectively on a shop floor.
| KPI | Green | Amber | Red | Escalation Action |
| OEE | ≥75% | 60–74% | <60% | Plant manager review. Root cause analysis within 48 hours. |
| First Pass Yield | ≥95% | 90–94% | <90% | Quality hold. Inspect last 2 hours of production. |
| OTIF | ≥95% | 88–94% | <88% | Customer communication. Expediting protocol activated. |
| Unplanned Downtime | <5% of planned hours | 5–10% | >10% | Maintenance escalation. Predictive diagnostics review. |
| Scrap Rate | <2% | 2–4% | >4% | Stop-and-fix protocol. Material and process audit. |
The three lines model maps directly onto this structure: the first line (production supervisors) monitors daily metrics and responds to amber triggers.
The second line (operations excellence or risk function) sets thresholds, validates data integrity, and conducts periodic risk assessments of the KPI program itself. The third line (internal audit) provides independent assurance that KPIs are measured accurately and that escalation protocols are followed.
Figure 5: Most manufacturing plants cluster at 55–60% OEE, with massive upside potential.
Building Momentum: Weeks 1 Through 12
Knowing which key production indicators to track is step one. Building the measurement infrastructure, training the team, and creating the governance structure to act on the data is where programs succeed or stall.
This phased approach applies the risk management lifecycle—identify, analyze, evaluate, treat, monitor—to your KPI implementation.
| Phase | Actions | Deliverables | Success Metrics |
| Weeks 1–4: Foundation | Audit current measurement capability. Identify data gaps for the five core KPIs. Install automated data capture where manual logging exists. Define green/amber/red thresholds for each KPI. | KPI measurement capability assessment. Data capture improvement plan. Threshold matrix approved by plant leadership. | All five core KPIs have defined data sources. Thresholds documented and communicated to all shift leads. |
| Weeks 5–8: Calibrate & Train | Run parallel measurement for 4 weeks to validate data accuracy. Train shift supervisors on KPI dashboards, threshold interpretation, and escalation protocol. Conduct first OEE deep-dive analysis. | Training completion records. Validated baseline for each KPI. First OEE loss waterfall analysis. | 100% of shift leads trained. Baseline OEE established with +/- 2% confidence. Top 3 loss categories identified. |
| Weeks 9–12: Act & Improve | Launch targeted improvement projects for the top 3 loss categories. Implement daily KPI standup meetings (15 min max). Conduct first monthly management review using the KRI escalation framework. | Improvement project charters (SMART objectives). Monthly KPI report with trend analysis. First management review minutes. | At least one loss category reduced by 10%+. Daily standups running with 90%+ attendance. Management review completed on schedule. |
Where Programs Stall — And How to Unstick Them
| Trap | Root Cause | Fix |
| Tracking 30+ KPIs | Fear of missing something. Every department adds their metrics. | Ruthlessly prioritize. The five core KPIs cover 80%+ of production risk. Add secondary metrics only for specific projects. |
| Manual data collection | Capital avoidance. “We’ll upgrade the system next year.” | Start with the simplest digital capture: a tablet at each workstation logging downtime reasons. Even basic digitization improves accuracy 3–5x. |
| KPIs as punishment tools | Metrics used to blame operators rather than improve processes. | Publicly celebrate improvement trends. Investigate red flags as systemic issues, not individual failures. Deming was right: 94% of problems are system problems. |
| No threshold structure | KPIs reported without green/amber/red context. Everything looks the same. | Apply the KRI framework above. Without thresholds, a KPI is just a number. With thresholds, it becomes a risk control. |
| Annual reviews only | KPIs reviewed quarterly or annually. Production problems happen hourly. | Daily standup (5–15 min) for operational KPIs. Weekly plant review for tactical KPIs. Monthly management review for strategic KPIs. |
| Ignoring leading indicators | Over-reliance on lagging metrics like scrap rate and customer complaints. | Balance leading indicators (planned maintenance %, WIP trend, changeover time) with lagging indicators. Lead with the metrics you can still act on. |
Emerging Threats Your KPI Program Isn’t Ready For
1. Cybersecurity-driven downtime. Siemens’ 2024 report found that manufacturing experiences more cybersecurity-related downtime than most industries, with revenue losses of $58 million attributable to cyber incidents.
As operational technology (OT) networks become more connected, manufacturers need IT risk management KPIs alongside production KPIs. Mean time to detect (MTTD) and mean time to respond (MTTR) for cyber events should sit on the same dashboard as OEE.
2. AI-driven predictive maintenance reshaping baselines. Predictive maintenance tools are reducing unplanned downtime by up to 50% at early adopters. As these tools become standard, the baseline for “good” OEE will shift upward. Plants still relying on reactive maintenance will fall further behind. ERM technology convergence with shop-floor IoT is the next frontier.
3. Supply chain volatility as a permanent KPI category. Material shortages contributed 18.4% of all efficiency losses in the Godlan study. Post-pandemic, post-trade-war, this is not a temporary disruption—it’s a structural feature of global manufacturing. Key production indicators now need a supply chain risk layer: supplier OTIF, raw material buffer days, and single-source dependency ratios all belong on the expanded KPI dashboard.
Build your production risk management system. Explore riskpublishing.com for practitioner-grade frameworks on risk assessment, KRI dashboards, business impact analysis, and operational resilience. Production excellence and risk management are two sides of the same coin.
References
1. Siemens. The True Cost of Downtime (2022, 2024) — Fortune Global 500 manufacturers lose 11% of annual revenue ($1.4 trillion) to unplanned downtime.
2. Godlan. OEE Benchmarks by Manufacturing Industry Vertical (2025) — 1,470+ discrete manufacturing operations analyzed. Medical devices highest at 78.2% OEE.
3. Evocon. World-Class OEE: Industry Benchmarks from 50+ Countries (2024) — 3,500+ machines; global average 55–60% OEE; ~3% achieve world-class.
4. SCW.ai. World-Class OEE in Pharma: A Benchmarking Analysis (2025) — Pharma OEE averages 35%. Moving to 60% could yield $14–$16M annually per plant.
5. ISM. The Monthly Metric: Unscheduled Downtime (2024) — Tesla Berlin incident; ABB reliability survey of 3,200+ plant maintenance leaders.
6. Aberdeen/ServiceMax. Cost of Downtime in Industrial Manufacturing — Average $260,000/hour; $2M per incident. 800 hours annual unplanned downtime.
7. LeanProduction.com / OEE.com. Overall Equipment Effectiveness Guide — World-class 85%, typical discrete manufacturing 60%, beginning trackers 40%.
8. Fluke. How to Reach Manufacturing Excellence with OEE Benchmarks (2025) — Practical OEE calculation walkthrough. 85% benchmark requires 95%+ on each sub-metric.
9. MDCplus. What Is a Good OEE Score? Industry Benchmarks & Case Studies — Industry-specific OEE data; Pfizer >90% on specific lines post-re-engineering.
10. Symestic. OEE Benchmarks: Facts, Realistic Values, and Practical Limitations — Cross-study compilation of OEE benchmarks with methodological context.
11. IDS-INDATA. The Real Cost of Downtime: Sector-by-Sector Breakdown (2025) — UK/EU manufacturers projected to lose £80 billion to downtime in 2025.
12. Erwood Group. True Costs of Downtime in 2025: Deep Dive by Business Size — Gartner and Siemens data by enterprise size and sector.
13. AccouNovation. Essential Manufacturing KPIs and Metrics for 2025 — Comprehensive KPI breakdown with OEE, cycle time, inventory turns, and OTD. 14. Sockeye. Unplanned Manufacturing Downtime: Cost, Causes & Prevention — Siemens 2023 data: $40K–$2M/hour downtime costs. M
Chris Ekai is a Risk Management expert with over 10 years of experience in the field. He has a Master’s(MSc) degree in Risk Management from University of Portsmouth and is a CPA and Finance professional. He currently works as a Content Manager at Risk Publishing, writing about Enterprise Risk Management, Business Continuity Management and Project Management.