MES & Shop FloorFree referencePDF, Excel, Word export

Production Scheduling KPI Formula Reference (Schedule Adherence, Utilization, Flow Time)

Scheduling KPIs sound simple until two people calculate them differently. This table lists the standard formulas with their units and the usual misreadings, so a scheduler can agree one definition and defend it.

Download:

KPI	Formula	Unit	What it tells you	Common misreading
Schedule adherence (by quantity)	Units produced as planned / units planned x 100	%	Did the floor make what was scheduled	Overproduction can hide a miss on other items
Schedule adherence (by job)	Jobs completed in the planned window / jobs planned x 100	%	How reliable the plan is job by job	A job one hour late counts the same as one three weeks late
Plan attainment	Planned hours earned / planned hours scheduled x 100	%	Weighted view of attainment by labor or machine hours	Depends on credible standard hours
On time delivery (OTD)	Orders shipped on or before the promise date / orders shipped x 100	%	Customer facing delivery performance	Date definition: promise date vs request date changes the result
On time in full (OTIF)	Orders on time and complete / orders due x 100	%	Delivery with correct quantity	Partial shipments need a clear tolerance rule
Capacity utilization	Actual output hours / available hours x 100	%	How much of available time was used	High utilization on a non-bottleneck often means overproduction
Load percent	Scheduled hours / available capacity hours x 100	%	Forward looking load vs capacity for a period	Over 100% means the plan cannot be met without action
Efficiency	Earned standard hours / actual hours worked x 100	%	Labor or machine performance vs standard	Standards that are stale make it meaningless
OEE	Availability x performance x quality	%	Equipment effectiveness during planned production time	Excludes planned downtime, so it is not TEEP
TEEP	OEE x utilization of calendar time	%	Effectiveness against all calendar time	Rarely comparable between plants with different shift patterns
Availability	Run time / planned production time	ratio	Loss from stops and breakdowns	Short stops often go unrecorded
Performance rate	(Ideal cycle time x total count) / run time	ratio	Loss from slow cycles	Wrong ideal cycle time distorts the result
Quality rate	Good count / total count	ratio	Loss from scrap and rework	Rework counted as good hides the loss
Takt time	Available production time / customer demand	time per unit	Pace needed to meet demand	Uses demand rate, not the production rate achieved
Cycle time	Time from start of one unit to start of the next unit at a station	time per unit	Pace of a process step	Not the same as lead time
Manufacturing lead time	Queue + setup + run + wait + move time	days	Total elapsed time through the routing	Queue time is usually the largest part and the least measured
Work in process (Little's law)	Throughput rate x average flow time	units	Links WIP, flow time and output	Throughput and flow time must use the same time unit
Average flow time	WIP / throughput rate	days	Estimated time a unit spends in the system	Assumes a stable system, not a surge period
Flow efficiency	Value adding time / total lead time x 100	%	Share of lead time spent actually working	Often below 10% in job shops, which is normal to see
Setup time ratio	Total setup time / (setup + run time) x 100	%	Share of capacity lost to changeover	Look at the bottleneck first, not the plant average
Backlog in weeks	Open order hours / weekly available capacity hours	weeks	How long the order book would take to clear	Ignores mix: backlog can sit on one resource
Past due backlog	Open operations with due date before today, count or hours	count or hours	Size of the late queue	Counting only jobs hides the hours
Schedule stability	Operations unchanged inside the freeze window / operations scheduled in the window x 100	%	How often the plan is reshuffled	Low stability often comes from late material, not from planners
Changeover frequency	Number of setups / scheduled production days	setups per day	Effect of batching and sequencing choices	Fewer setups can mean larger WIP
First pass yield	Units passing first time / units started x 100	%	Process quality at a step	Rolled throughput yield is the product of step yields

Formulas are the common textbook and shop floor definitions. Agree an exact definition with your management before you publish numbers, especially for adherence and on time delivery.

Want these KPIs computed from your ERP automatically?

Tell us which ERP you run and which KPIs you report by hand. Optional, the reference stays free.

Optional. The tool and every export work without this. No spam.

How to use this reference

Search for the KPI name, then copy the formula into your report specification or BI model. Check the unit column before you chart anything: Little's law only works when throughput and flow time use the same time unit, and takt time is a time per unit, not a rate. Export the filtered rows to Excel to build a KPI dictionary your team can sign off. A one page dictionary prevents most arguments about whose number is right.

Adherence versus attainment versus OTD

Schedule adherence looks inward: did the floor follow the plan. On time delivery looks outward: did the customer get the order by the promise date. A plant can have high adherence and poor delivery if the plan was wrong, and the reverse if the team recovers through overtime. Track both, and add schedule stability to see how often the plan changes after it was frozen. Together the three show whether problems come from planning, execution or demand.

Worked example with Little's law

Suppose a cell completes 40 units a day and holds 200 units of WIP. Average flow time is WIP divided by throughput, so 200 / 40 equals 5 days. If you cut WIP to 120 with the same throughput, flow time falls to 3 days. This is why WIP caps such as CONWIP or kanban shorten lead time without any speed change. If throughput also falls after the cut, you removed WIP from the bottleneck, which is the mistake to avoid.

Common mistakes

Reporting average utilization across all resources hides the bottleneck, whose utilization is the only one that limits output. Measuring adherence against a plan that was rewritten daily gives a flattering number. Mixing planned and unplanned downtime inside availability makes OEE look worse than the equipment really is. Define the plan snapshot date, the time basis and the exclusions in writing, and apply them the same way every period.

Frequently Asked Questions

How do I calculate schedule adherence?

Divide units, jobs or hours completed as planned by those planned in the same period and multiply by 100. State whether you measure by quantity, by job or by earned hours, because the three give different answers for the same week.

What is the difference between OEE and TEEP?

OEE measures effectiveness during planned production time using availability, performance and quality. TEEP multiplies OEE by the share of calendar time you plan to run, so it shows the loss from idle shifts as well.

What is a good capacity utilization target?

It depends on variability. A bottleneck is often planned near 85 to 95 percent, while resources with high variability need more spare time. Running every resource at 100 percent normally lengthens queues and lead times sharply.

How does Little's law help a scheduler?

It links WIP, throughput and flow time: WIP equals throughput times flow time. If you need shorter lead time without more output, you reduce WIP through release control, which is the idea behind CONWIP and kanban.

If your ERP holds the data but the KPIs live in spreadsheets, we can build these measures over live ERP tables with the query visible and read-only access.

Talk to an Expert Explore More Free Tools

Related Tools

MES & Shop Floor

More MES & Shop Floor tools

See all

Back to All Tools