Factory managers have more production data available than ever. But collecting more data does not automatically result in better decisions.
The real value comes from tracking a focused set of production metrics that show where capacity is being lost, why output is falling behind and which improvements will have the greatest operational impact.
Production analytics platforms can bring together machine activity, downtime, output, quality, maintenance and energy data. This gives factory teams a more complete view of performance than spreadsheets, paper records or isolated system reports.
But which production metrics should you prioritise?
This guide covers 10 of the most useful production metrics for factory managers, including how to calculate them, what they reveal and how to use them to improve performance.
For a broader overview covering productivity, quality, financial and customer measures, read our complete guide to key manufacturing metrics.
| Production Metric |
What it helps you understand |
|
Overall Equipment Effectiveness |
How effectively equipment is producing good parts during planned production time |
|
Machine utilisation |
How much of the available production time equipment is actively being used |
|
Unplanned downtime |
How much production time is being lost unexpectedly, and why |
|
MTBF and MTTR |
How reliable equipment is and how quickly failures are resolved |
|
First-pass yield |
How much output meets requirements without rework |
|
Scrap and rework rate |
How much capacity, material and labour are being lost to quality problems |
|
Throughput |
How quickly the factory produces acceptable output |
|
Cycle time against takt time |
Whether production is running quickly and consistently enough to meet demand |
|
Schedule attainment |
Whether the factory is producing what was planned, when it was planned |
|
Energy use per good unit |
How efficiently energy is converted into saleable output |
Machine utilisation measures the percentage of a defined period during which a machine is actively producing.
A typical calculation is:
Machine utilisation = Productive machine time ÷ Available production time × 100
Unlike OEE, utilisation can also reveal how much of the factory’s wider available capacity is being used. This is particularly valuable for manufacturers that need to increase output, reduce lead times or decide whether new equipment is really required.
Low utilisation does not necessarily mean a poorly performing machine. The cause could be insufficient demand, gaps in the production schedule, operator availability, material shortages, lengthy setups or an upstream constraint.
How to use machine utilisation
Segment utilisation data by machine, shift, job and product. This can reveal:
Machines with consistent unused capacity
Differences in performance between shifts
Jobs that repeatedly create long idle periods
Equipment that appears fully loaded but produces less than expected
Opportunities to move work away from bottleneck machines
Make sure the denominator is clearly defined. Comparing productive time with a scheduled shift answers a different question from comparing it with every hour in the week.
Overall Equipment Effectiveness, or OEE, measures how effectively production equipment is being used during planned production time.
It combines three components:
Availability: the percentage of planned production time during which the equipment was running
Performance: the speed at which the equipment ran compared with its ideal production rate
Quality: the proportion of output that met the required standard
The formula is:
OEE = Availability × Performance × Quality
OEE helps factory managers see beyond a simple running-or-stopped machine status. A line may be operating for most of a shift but still lose significant capacity through reduced running speeds, short stops or rejected parts.
How to use OEE
The overall percentage is only a starting point. Break OEE down by machine, line, shift, product or job to identify which component is creating the largest loss.
For example:
Low availability directs attention towards downtime and changeovers.
Low performance suggests slow cycles, micro-stops or speed losses.
Low quality points towards rejects, process instability or rework.
Avoid treating OEE as a score that every machine must maximise. It is most useful when it helps teams locate and quantify specific production losses.
Read our guide to what OEE is and how manufacturers can use it.
Unplanned downtime is the production time lost to unexpected stoppages, including equipment faults, tooling problems, material shortages and other unscheduled interruptions.
It can be expressed as a total duration or as a percentage:
Unplanned downtime rate = Unplanned downtime ÷ Planned production time × 100
Downtime affects more than the stopped machine. It can reduce throughput, disrupt schedules, create overtime and put customer delivery dates at risk.
However, recording the duration of downtime alone only tells you how much time was lost. To improve performance, you also need to know why it was lost.
How to use downtime data
Record downtime against consistent reason codes and analyse it by:
Machine or production line
Cause
Shift
Product or job
Duration
Frequency
Time of day
Use Pareto analysis to distinguish between frequent short stops and less frequent but longer disruptions. One recurring five-minute delay may consume more capacity over a month than a highly visible two-hour breakdown.
Focus first on causes that create the greatest cumulative loss, rather than simply reacting to the most recent incident.
Learn more about why understanding machine downtime is essential.
Mean Time Between Failures and Mean Time to Repair provide two complementary views of equipment reliability.
Mean Time Between Failures
MTBF estimates how long a repairable asset operates, on average, before it fails.
MTBF = Total operating time ÷ Number of failures
A declining MTBF can indicate that a reliability problem is developing, even if the machine is currently meeting its production target.
Mean Time to Repair
MTTR measures the average time required to restore equipment following a failure.
MTTR = Total repair time ÷ Number of repairs
High MTTR can result from difficult diagnosis, unavailable spares, missing documentation, limited technical coverage or delays in accessing the machine.
How to use MTBF and MTTR
Track both measures together. MTBF tells you how often failures occur, while MTTR shows how effectively the business responds when they happen.
Compare the figures by asset, fault type and time period. This can help maintenance and operations teams decide whether to:
Change preventive-maintenance intervals
Hold different critical spares
Improve fault documentation
Provide additional technical training
Investigate a recurring component failure
Standardise a successful maintenance approach across similar assets
Use these metrics selectively. They are most valuable for critical and repairable equipment, particularly machines whose failure would constrain the wider production process.
First-pass yield measures the percentage of units that meet quality requirements the first time, without rework, repair or additional processing.
First-pass yield = Good units produced first time ÷ Total units entering the process × 100
This metric connects quality directly with productivity. Rework consumes machine time, labour, materials and energy, even when the affected units are eventually accepted.
A factory can appear to be hitting its output target while losing considerable capacity behind the scenes through repeated correction and inspection.
How to use first-pass yield
Analyse first-pass yield by:
Product or SKU
Machine
Process stage
Shift
Material or supplier batch
Tool or machine setting
Look for changes in first-pass yield following a new material batch, setup adjustment, tool change or maintenance intervention. Connecting quality results with process conditions can help teams address variation before it produces significant scrap.
Do not confuse first-pass yield with final yield. Final yield may count a reworked item as acceptable, whereas first-pass yield exposes the additional effort required to make it acceptable.
Scrap and rework are related production losses, but they should be recorded separately.
Scrap is material or output that cannot be recovered:
Scrap rate = Scrapped units ÷ Total units produced × 100
Rework is output requiring additional processing before it can be accepted:
Rework rate = Reworked units ÷ Total units produced × 100
Both measures help reveal the true cost of quality problems. Scrap wastes material and embedded production cost, while rework also uses capacity that could otherwise be producing new saleable output.
How to use scrap and rework data
Categorise losses so teams can distinguish between:
Process-related defects
Material problems
Setup and changeover errors
Equipment-condition issues
Tool wear
Operator or training-related variation
Design or specification problems
Where possible, measure both the quantity and cost of the loss. Ten rejected high-value components may have a greater commercial impact than hundreds of inexpensive items.
The goal is not simply to report waste at the end of the month. Production teams need sufficiently timely information to intervene before a recurring problem affects the rest of the batch.
Throughput is the rate at which a process, line or factory produces acceptable output.
Throughput = Good units produced ÷ Production time
Using good output prevents rejects, incomplete products and work-in-progress from inflating the result.
Throughput is particularly useful because it reflects the combined effect of availability, production speed, quality and flow. It provides a direct indication of whether the factory is converting its capacity into customer-ready products.
How to use throughput
Compare actual throughput with the required or planned rate during the shift, not only after production has finished.
If throughput falls behind, investigate whether the constraint is:
Machine downtime
Reduced operating speed
Quality losses
Labour availability
Material shortages
An upstream process
An accumulation of work-in-progress
Avoid using factory-wide averages alone. Overall throughput may conceal a bottleneck at one process stage that limits the output of the entire production system.
Cycle time is the actual time taken to produce one unit or complete a process step.
Average cycle time = Production time ÷ Units produced
Takt time is the pace required to satisfy customer demand:
Takt time = Available production time ÷ Customer demand
Comparing the two answers an important operational question: can the process consistently produce quickly enough to meet demand?
If cycle time regularly exceeds takt time, a backlog will develop unless capacity, staffing, scheduling or process performance changes.
How to use cycle-time data
Track the distribution and variation, not just the average.
Two machines could have the same average cycle time while behaving very differently. One might run consistently, while the other alternates between very fast and very slow cycles. The second process will usually be harder to schedule and may indicate inconsistent material flow, tool wear, operator differences or intermittent interruptions.
Compare cycle times for equivalent products and operating conditions. Mixing different jobs into one average can create a misleading performance figure.
Schedule attainment measures how closely actual production matched the planned schedule.
A simple volume-based calculation is:
Schedule attainment = Scheduled production completed ÷ Planned production × 100
The metric shows whether the factory produced the right products, in the required quantities, within the planned period.
This distinction matters. A factory can achieve a high overall production count while still missing priority orders or producing the wrong product mix.
How to use schedule attainment
When the schedule is missed, record the reasons alongside the result. Common causes include:
Equipment downtime
Material shortages
Quality holds
Labour constraints
Changeover delays
Inaccurate standard times
Unplanned priority changes
Unrealistic production plans
This helps operations and planning teams separate execution problems from weaknesses in the original plan.
In factories with a varied product mix, consider calculating schedule attainment using standard production hours or completed orders rather than simple unit counts. Otherwise, a large quantity of easy-to-produce items may conceal missed lower-volume, time-intensive work.
Although, not always in a managers list of production metrics to track, recent rises in energy costs, changes to regulation and a greater focus on sustainability, energy use is our number 10 metric to track.
Energy use per good unit measures how efficiently the production process converts energy into acceptable output.
Energy per good unit = Total production energy consumed ÷ Good units produced
A total monthly electricity figure tells you what the factory consumed, but it does not explain how efficiently that energy supported production. Normalising consumption against good output makes comparisons more meaningful.
How to use energy-production data
Compare energy per good unit by:
Machine
Product
Shift
Production line
Operating condition
Time period
This can reveal idle energy consumption, inefficient equipment, air leaks, unsuitable process settings or products that require disproportionately high energy use.
Keep the measurement boundary consistent. Decide whether the calculation includes only machine energy or also supporting systems such as extraction, compressed air and cooling.
Connecting energy data with production, downtime and quality information can also identify situations in which rising consumption is an early indicator of deteriorating machine or process performance.
A production analytics platform may need data from several sources to provide a complete view.
Data source |
Machine-level energy monitoring |
| Machine monitoring system | Machine status, utilisation, downtime and cycle times |
| Production counting
or quality system |
Throughput, first-pass yield, scrap and rework |
| CMMS or maintenance records | Failure history, MTBF and MTTR |
| ERP, MES or production-planning system | Production schedules and schedule attainment |
| Machine-level energy monitoring | Energy consumption per machine, job or unit |
Before combining data, agree consistent definitions. For example, “productive”, “available”, “failure” and “good unit” must mean the same thing across shifts, departments and sites.
A production dashboard should help people decide what to do next, not simply describe what happened.
Every metric should have:
A clearly defined purpose
An agreed calculation
A reliable data source
An accountable owner
A target or expected range
A regular review rhythm
A defined response when performance moves outside that range
Teams should also review metrics together rather than in isolation.
For example, rising throughput might initially look positive. But if first-pass yield is falling and energy per good unit is increasing, the additional output may be creating higher costs and more rework.
Similarly, a high utilisation figure is not automatically desirable if machines are producing unnecessary inventory or if constant loading prevents essential maintenance.
The most useful production metrics balance output, time, quality, reliability and cost.
Production metrics are valuable when they make losses visible and help teams prioritise action.
Start with a small number of measures linked to a genuine factory objective. Establish a reliable baseline, investigate the underlying causes of lost performance and use the same data to verify whether changes deliver a sustained result.
If you are still deciding which measures are appropriate for your operation, explore our complete guide to key manufacturing metrics. It covers a wider range of productivity, quality and financial KPIs, together with guidance on implementing them successfully.
FourJaw’s production tracking software helps manufacturers capture and analyse real-time machine utilisation, downtime, production and energy data. By making factory performance visible, teams can identify hidden capacity, focus improvement activity and make decisions using objective production data.