Operations KPIs: Throughput, Cycle Time and On-Time Delivery
Operations KPIs explained: throughput, cycle time, lead time, on-time delivery, utilisation and first-pass yield, with formulas and a worked rand example.
· 4 min read · Summarix team
Operations KPIs measure how well you turn inputs into delivered work. The essential ones are throughput (how much you complete), cycle time (how long each unit takes), on-time delivery (whether you keep promises), utilisation (how busy your resources are) and first-pass yield (how much is right first time). They apply to a factory, a workshop, a warehouse or a service team.
The core operations KPIs
| KPI | Formula |
|---|---|
| Throughput | Units (orders, jobs, tickets) completed ÷ time period |
| Cycle time | Average time from work started to work finished |
| Lead time | Average time from customer order to delivery |
| Work in progress (WIP) | Units started but not finished at a point in time |
| On-time delivery (OTD) | Orders delivered by promised date ÷ total orders × 100 |
| OTIF (on time, in full) | Orders delivered on time and complete ÷ total orders × 100 |
| Utilisation | Productive hours ÷ available hours × 100 |
| First-pass yield | Units right first time ÷ units started × 100 |
| Cost per unit | Total operating cost ÷ units completed |
Little’s Law: how throughput, WIP and cycle time connect
These three are linked by a simple relationship called Little’s Law: average cycle time = average WIP ÷ average throughput. If a workshop has 60 jobs in progress and finishes 20 a day, each job takes about 3 days. The practical lesson: if you want faster delivery without adding capacity, reduce work in progress. Starting more jobs than you can finish just makes every job slower.
Worked example: a joinery workshop
Say your joinery business completed 180 orders in a 20-working-day month. On average there were 54 orders in production at any time. 153 orders were delivered by the promised date, 144 of them complete, and 162 passed quality inspection first time. Operating costs for the month were R540,000. Your six craftsmen had 960 available hours and logged 768 hours on jobs.
- Throughput: 180 ÷ 20 = 9 orders a day
- Cycle time (Little’s Law): 54 ÷ 9 = 6 working days
- On-time delivery: 153 ÷ 180 = 85%
- OTIF: 144 ÷ 180 = 80%
- First-pass yield: 162 ÷ 180 = 90%
- Utilisation: 768 ÷ 960 = 80%
- Cost per order: R540,000 ÷ 180 = R3,000
The story: 18 orders a month (10%) need rework, and rework is often the hidden cause of late deliveries. If each rework takes a day of a craftsman’s time, improving first-pass yield to 95% frees around 9 working days a month, which could lift throughput and on-time delivery at the same time. That is a better first move than hiring. Cost per order tells the same story in rand: with R540,000 of monthly costs, lifting throughput from 180 to 189 orders on the same cost base would bring cost per order down from R3,000 to about R2,857.
On-time delivery vs OTIF
On-time delivery alone can hide short shipments. OTIF only counts an order as successful if it arrived on time and complete, which is how most customers see it. Agree what ‘on time’ means (the promised date, the requested date, or within a window) and apply it consistently, otherwise the KPI becomes a debate rather than a measure.
Cycle time vs lead time
Cycle time starts when work begins; lead time starts when the customer orders. The gap is queue time: orders waiting for materials, scheduling or a free workstation. In many operations the queue is much longer than the work itself. Measuring both shows you whether to speed up the work or reduce the waiting.
Upload your job or order log and Summarix calculates throughput, cycle time and on-time delivery, flags anomalies and writes up the trends.
Free plan: 5 AI reports a month, no card needed.
Running a weekly operations review
- Throughput and WIP this week vs the previous four weeks.
- OTIF, with the reasons for each late or incomplete order grouped into categories.
- First-pass yield and the top rework causes.
- Utilisation by team or machine, looking for bottlenecks rather than idle time.
- One improvement to trial next week, and how you will know if it worked.
Look for sudden changes as well as trends: a spike in cycle time or drop in yield often points to a specific supplier, shift or machine. Our guide to spotting anomalies in business data covers simple methods. Summarix can take a job log from a spreadsheet, database or tool like Jira or monday.com and produce these KPIs automatically, so the review starts with the numbers already prepared.
Conclusion
Operations KPIs work best as a connected set. Throughput, WIP and cycle time explain speed; OTIF shows what customers experience; first-pass yield and utilisation explain why. Track them weekly, fix rework before adding capacity, and keep enough slack to keep your promises. For choosing which of these matter most for your business, see how to choose the right KPIs, and for a broader starter set, 15 KPIs every small business should track.
Frequently asked questions
What is the difference between throughput and cycle time?
Throughput is how many units you complete in a period. Cycle time is how long each unit takes from start to finish. They are linked through work in progress by Little’s Law.
How do you calculate OTIF?
Divide the number of orders delivered both on time and complete by the total number of orders, then multiply by 100.
What is a good utilisation rate?
Very high utilisation usually leads to long queues and late deliveries. Many operations aim for a level that leaves some slack for disruptions, and the right figure depends on how variable your work is.
What is first-pass yield?
The percentage of units that pass quality checks the first time without rework. It is a strong indicator of process quality and hidden cost.