Manufacturing Lead Time: It's Mostly Waiting, Not Working

A steel part is machined for 40 minutes. It spends 9 days in the factory. Nobody was lazy, no machine broke, and the order still went out "on schedule". Where did the other 8 days and 23 hours go?

Manufacturing lead time is the total time from releasing an order to production until the finished product is ready to ship. It includes every minute the part is being worked on, and every minute it's waiting. In most plants, the waiting wins by a wide margin.

The five parts of manufacturing lead time

Every hour of lead time falls into one of five buckets:

• Queue time: waiting in front of a machine for the jobs ahead to finish.
• Setup time: the machine being changed over for this job.
• Run (processing) time: the part actually being worked on.
• Wait time: after processing, waiting for the rest of the batch to finish before moving on.
• Move time: transport between work centers.

Only run time changes the part. The other four just add days.

A quick word on terms: cycle time usually means the time per unit at one station (40 minutes at the lathe here). Lead time is the whole journey. Mixing them up is how a "40-minute part" gets promised in two days.

A worked example: 9 days, broken down

The part goes through three operations: sawing, turning and deburring/inspection. It's made in batches of 50. Run time per part is 40 minutes of turning plus 20 minutes across the other two operations, so 1 hour in total. Nine calendar days, with the plant running around the clock, is 216 hours.

• Run: 1 hour.
• Setup: 2.5 hours across the three operations.
• Move: 6 hours, mostly waiting for a forklift.
• Wait: at turning, the part waits while its 49 batch-mates are machined: 49 × 40 min = 1,960 min, or 32.7 hours. At the other two operations: 49 × 20 min = 980 min, or 16.3 hours. Total: 49 hours.
• Queue: everything else. 216 − 1 − 2.5 − 6 − 49 = 157.5 hours, about 6.6 days.

Queue is 73% of the lead time. Batch waiting is another 23%. Run time is 1 hour out of 216, less than half of one percent.

Think of a one-hour flight. Door to door, the trip takes five: driving, parking, security, boarding, taxiing, baggage claim. Nobody tries to shorten that trip by buying a faster plane.

Myth: "A faster machine shortens lead time"

Say you buy a lathe that's 20% faster. Turning drops from 40 to 32 minutes per part. For one part, you save 8 minutes of run time, and the batch wait at turning drops from 32.7 to 26.1 hours. That's worth having. But the 157.5 hours of queue don't move at all, because the queue is set by how much work is waiting, not by how fast one machine runs. Unless that lathe is the bottleneck, the part still takes about 8.7 days.

Why the ERP's fixed lead time drifts

Most ERPs store lead time as a fixed number per item, often typed in years ago: "10 days". But queue time isn't fixed. It depends on how much work is on the floor right now. Little's law says lead time = WIP ÷ throughput. If about 330 parts sit ahead of the line and it finishes 50 a day, the queue is 330 ÷ 50 = 6.6 days. Add 150 more parts of work and it becomes 480 ÷ 50 = 9.6 days. The ERP still says 10, whether the real number is 6 or 12. That's why ERP thinks every order takes 10 days. Work in progress and Little's law explains the formula in detail.

What actually shortens lead time

1. Cap the work on the floor. Queue time is WIP divided by throughput. Release less and the queue shrinks, without losing output, as long as the bottleneck always has work.

2. Cut batch sizes, or split them for transfer. Halve the batch from 50 to 25 and the wait at turning becomes 24 × 40 min = 16 hours instead of 32.7. Across all three operations, wait drops from 49 to 24 hours, a full day saved. The cost is twice as many setups, so it pays where setups are short. Even simpler: move the first 10 finished parts to the next operation instead of waiting for all 50.

3. Protect the bottleneck. An hour lost at the bottleneck adds queue for everyone behind it. An hour lost elsewhere often costs nothing.

4. Schedule on real capacity. Production scheduling with finite capacity gives each order a date from the actual load, not a fixed number from the item master.

Shorter and steadier lead times pay off twice: customers get faster answers, and you need less buffer stock, because safety stock grows with lead time and its variability.

How this looks in factory.online

In factory.online, finite-capacity scheduling calculates each order's dates from what's already loaded on each resource, so the planner sees the part's real finish date, not a fixed 10 days. Output confirmations from the floor show when each operation actually ran, which makes the gap between 1 hour of run time and 9 days of lead time visible order by order.

If your quoted lead times are mostly padding, book a demo and we'll break one of your orders down together.