Lead Time

Cutting Manufacturing Lead Time Systematically

In most plants a part spends over 90 percent of its lead time waiting. Map the five components, attack queue time first, and hold the gains with a weekly rhythm.

Lead time is a pricing weapon and an inventory tax at the same time. Quote 6 weeks against a competitor's 3 and you lose the order without ever discussing price; carry a 6 week internal lead time and Little's Law forces you to hold twice the WIP of a 3 week plant at equal output. A plant shipping $80,000 a day at 12 days of internal lead time holds $960,000 in WIP by arithmetic. Cut lead time to 5 days and $560,000 of cash walks off the floor and into the bank.

Measure it in five buckets: queue, setup, run, inspection, and move. Take a real order through the shop with timestamps. Typical result on a 12 day lead time: queue 10.2 days, setup 0.3, run 0.8, inspection 0.5, move 0.2. The part is being worked on 7 percent of the time and waiting 93 percent. That ratio, often called flow efficiency, runs 5 to 10 percent in most job shops and batch plants. The Lead Time calculator adds the five components in minutes; the revelation is always the queue line.

Because queue dominates, attack it first and leave the cycle time projects for later. Queue time is driven by three things: loading, batch size, and variability. Above 85 percent machine loading, queues grow nonlinearly; a work center at 95 percent loading can easily carry 5 to 10 times the queue of one at 80 percent. Batch size sets how long each order waits behind others: cutting batches from 2,000 to 500 units at a shared work center typically cuts queue time 40 to 60 percent by itself. Variability, from breakdowns and hot-list chaos, does the rest.

Setup reduction is the enabler that makes small batches affordable. A 90 minute changeover taken to 15 minutes through SMED lets you run one quarter the batch size at the same setup cost per unit, and lead time falls with the batch. Then fix release discipline: releasing work to the floor faster than the constraint can consume it does not make anything ship sooner, it just grows every queue. Gating release to constraint capacity, the core of any pull or CONWIP system, routinely cuts shop lead time 30 to 50 percent without touching a single machine.

Inspection and move time are smaller but cheap to fix. Inspection queues of 1 to 2 days are common where every lot waits for a single inspector; moving to operator self-check with audit sampling cuts that to hours. Move time hides in layout: a part traveling 1,500 feet across a plant in 8 handling steps picks up hours of waiting at each staging spot. Cellularizing a product family typically cuts travel 70 to 90 percent and removes the staging queues with it.

The failure modes are predictable. Expediting as a management system: every hot order that jumps a queue adds queue time to every other order, and plants with 30 percent of orders expedited have effectively randomized their lead times. Measuring only averages: a 12 day average with a standard deviation of 6 means quoting 18 to 24 days to be safe, so variance reduction is worth as much as mean reduction. And starting with a machine upgrade: speeding up run time attacks the 7 percent of lead time that was never the problem.

Run it as a cadence. Daily: the tier meeting reviews orders aging past their queue standard at each work center and clears blockers. Weekly: track lead time by value stream, flow efficiency, and percent of orders expedited; anything over 10 percent expedited triggers a release review. Monthly: re-walk one product family with timestamps and update the five-bucket map. World-class looks like flow efficiency above 25 percent, quoted lead times half the industry norm, expedites under 5 percent, and lead time quoted from measured data instead of hope. Plants that get there typically cut lead time 50 to 70 percent in 12 to 18 months, mostly from queues, not equipment.

Published 2026-07-02.