Title: How to Reduce Manufacturing Lead Time URL: https://zudoworks.com/manufacturing/guides/reduce-manufacturing-lead-time/ Description: How small factories cut manufacturing lead time: measure each job, cap work in progress, release only with material and protect the bottleneck. Free calculator. Updated: 2026-11-05 [Home](https://zudoworks.com/) [Manufacturing](https://zudoworks.com/manufacturing/) [Guides](https://zudoworks.com/manufacturing/guides/) Planning # How to Reduce Manufacturing Lead Time: A Practical Guide Most of a job’s lead time is spent waiting, not being made. Here is how a small factory finds the waiting and removes it, without buying new machines. ![Arunkumar V](https://zudoworks.com/Asset/img/team/arunkumar-v-avatar.webp) By [Arunkumar V](https://zudoworks.com/about/#leadership), founder and CTO of Zudo Works 5 November 2026 · 17 min read Who it is for : Factories of 10 to 250 people Key formula : Lead time = jobs open ÷ jobs finished per day Try it : Lead time calculator Plan : Eight weeks, first change in weeks 3 and 4 - [Overview](https://zudoworks.com/manufacturing/) - [Zoho implementation](https://zudoworks.com/manufacturing/zoho-implementation/) - [AI automation](https://zudoworks.com/manufacturing/ai-automation/) - [Workflows](https://zudoworks.com/manufacturing/workflows/) - [Guides](https://zudoworks.com/manufacturing/guides/) - [Blueprints](https://zudoworks.com/manufacturing/case-studies/) - [Software](https://zudoworks.com/manufacturing-software/) ![Cover: how to reduce manufacturing lead time in four steps, measure, limit work in progress, unblock material and protect the bottleneck](https://zudoworks.com/Asset/img/manufacturing/reduce-manufacturing-lead-time.webp) Measure, limit work in progress, unblock material and protect the bottleneck. On this page 1. [What manufacturing lead time is, and the three versions you should track](#what-lead-time-is) 2. [The one formula that governs lead time](#littles-law) 3. [Step one: measure lead time per job](#measure) 4. [Eight levers that shorten lead time](#levers) 5. [Seeing exactly where each job waits](#where-jobs-wait) 6. [What to measure every week](#measures) 7. [Common mistakes that make lead times longer](#mistakes) 8. [Where AI helps with lead time, and where it does not](#ai) 9. [Where ZMOS fits](#zmos) 10. [An eight-week plan](#plan) 11. [Frequently asked questions](#faq) In short To reduce manufacturing lead time in a small factory, cut the waiting rather than the work. Measure lead time per job, cap the number of jobs open on the floor, release only jobs that have their material, keep the bottleneck fed and never idle, shrink setups and transfer batches, and promise dates from the real load. By Little’s law, fewer open jobs at the same output means shorter lead times. Ask a customer what they want from a small manufacturer and two answers come back almost every time: the right part, and sooner. Lead time is the time from the moment an order is accepted to the moment it leaves the building, and for most job shops and small factories it is the number that wins or loses the next order. It is also the number most owners believe can only come down with more machines, more people or more overtime. That belief is usually wrong. In most small factories the work itself, the minutes a machine or a person actually spends on a job, is a small part of the lead time. The rest is waiting: for a decision, for material, for a busy machine, for the next step to notice the job is ready. This guide is for owners, works managers and operations leads in factories of roughly 10 to 250 people. It explains what lead time is made of, the one formula that governs it, eight practical levers, how to measure progress, and an eight-week plan. There is a calculator you can use with your own numbers as you read. ## What manufacturing lead time is, and the three versions you should track People use “lead time” for different things, which is one reason it is hard to improve. Agree three definitions inside the business and write them down: - **Order lead time:** from order accepted to goods dispatched. This is what the customer feels and what you quote. - **Floor lead time:** from job released to the floor to the last step completed. This is what production controls directly. - **Supplier lead time:** from purchase order placed to material received. This is what purchasing controls, and it sets the minimum for any job that needs bought-in material. The gap between order lead time and floor lead time is often a surprise. It is the time an accepted order spends being checked, priced, planned and waiting for release. In many small businesses that part alone takes days, and nobody owns it, because it sits between sales, the office and the floor. ![The five parts of manufacturing lead time: order entry and planning, waiting for material, queue before each step, setup and run, and move, inspect and dispatch](https://zudoworks.com/Asset/img/manufacturing/reduce-manufacturing-lead-time-anatomy.webp) Where the days go. Only one of the five rows is value-adding work; the others are waiting of one kind or another. Look at the rows above and ask which ones a customer is paying for. Only setup and run change the part. Everything else is waiting, moving or checking. Lean practitioners have long pointed out that in discrete manufacturing, waiting in queues is usually the largest part of lead time by far. You do not need to trust anyone’s percentage: the calculator further down lets you work out your own share in a few minutes. ## The one formula that governs lead time There is a simple relationship between three numbers every factory has: how many jobs are open on the floor, how many jobs the floor finishes per day, and how long a job takes to get through. It is called Little’s law, after John Little, who published the proof in the journal Operations Research in 1961: ![Little’s law: lead time equals work in progress divided by throughput. Example: 60 open jobs divided by 15 jobs a day gives 4 days](https://zudoworks.com/Asset/img/manufacturing/reduce-manufacturing-lead-time-littles-law.webp) Little’s law, proved by John Little in 1961, holds for long-run averages in a stable system. Example numbers only. In plain words: **average lead time equals the number of jobs open divided by the number finished per day**. If 60 jobs are open and the floor finishes 15 a day, the average job spends four days on the floor. The law holds for long-run averages in a reasonably stable system, which describes most factories over a month or two. It does not care what the jobs are, how they are scheduled or how variable they are. The key consequence The useful consequence is this. If output stays the same and you halve the number of jobs open, the average lead time halves. Nothing about the machines or the people has changed; there is simply less in the queue in front of every step. That is why the most effective lead time programmes in small factories start by controlling what is released, not by speeding up machines. The obvious worry is that fewer open jobs means less output. It does not, as long as the bottleneck never runs out of work. Output is set by the slowest step. Every job open beyond what keeps that step busy adds waiting time without adding a single part shipped. The skill is to keep just enough work in front of the bottleneck, and no more everywhere else. ## Step one: measure lead time per job You cannot reduce what you cannot see, and an average across the year hides everything useful. Measure lead time for every job, using dates you almost certainly already have: when the order was accepted, when the job was released, when each step was finished and when it was dispatched. Start with the last eight to twelve weeks of finished jobs. For each one, work out order lead time and floor lead time in working days. Then sort them. Three patterns appear in most small factories. A handful of jobs take far longer than the rest, usually because they waited for material or were started and parked. A group of jobs waited days before release. And jobs of the same product took very different times, which tells you the work itself is not what decides the lead time. Use the calculator below for a quick first view, without any spreadsheet work. Count the jobs open on the floor today, released but not finished, and the jobs you finish in a typical week. Then enter how many hands-on hours a typical job really needs across all its steps. ### Lead time calculator (Little’s law) Count the jobs open on your floor today and the jobs you finish in a typical week. Everything is calculated in your browser. Jobs open on the floor now Jobs finished per week Working days per week Hands-on hours per job (setup and run, all steps) Working hours per day Target jobs open on the floor Average floor lead time Hands-on time per job Share of lead time spent waiting Lead time at your target Formula: lead time = jobs open ÷ (jobs finished per week ÷ working days). It holds for averages over several weeks, while output stays steady. The target line assumes output does not fall, which is true as long as the bottleneck never runs out of work. With the example figures, the average job spends four days on the floor and needs under half a day of hands-on work, so roughly nine-tenths of its time is spent waiting. Your own numbers will differ, but if the share of waiting comes out above about 75%, the fastest gains will come from controlling work in progress and release, not from faster machines. Our answer on [how to track work in progress](https://zudoworks.com/answers/how-to-track-wip/) covers how to count open jobs reliably. ## Eight levers that shorten lead time There is no single fix. There is a short list of levers that work in almost every small factory, and the order matters: the first four are habits and rules that cost little, the last four need more effort and pay back once the first four are in place. ![Checklist of eight practical levers to reduce manufacturing lead time](https://zudoworks.com/Asset/img/manufacturing/reduce-manufacturing-lead-time-levers.webp) Eight levers, roughly in the order a small factory should pull them. The first four cost almost nothing. ### 1. Cap the number of jobs open on the floor Set a limit on how many jobs can be released and unfinished at the same time. A practical starting point is the number open today, reduced by a fifth to a quarter, then adjusted every two weeks. When the floor is at the limit, a new job is released only when another finishes. This feels uncomfortable for the first week, because the instinct is to start everything so it looks busy. Within two or three weeks the queues in front of each step shorten, jobs move faster, and the floor is calmer. A board that shows released and in-progress jobs as columns makes the limit visible to everyone. ![ZMOS Jobs board with work orders in Draft, Released, In progress, Completed and Closed columns, each card showing part, quantity, due date and progress, with overdue dates in red](https://zudoworks.com/Asset/img/manufacturing/zmos/zmos-jobs-board.webp) ZMOS screen The Jobs board in ZMOS. Counting the cards in the Released and In progress columns gives you work in progress at a glance, and overdue dates show in red. Sample data. ### 2. Release only jobs that have their material The release check A job that starts and then stops for a missing part is the worst kind of work in progress. It takes up a machine, it takes up floor space, it has to be tracked, and it still is not finished. A **release check** fixes this: before a job goes to the floor, confirm that every material on its bill is in stock or due before the step that needs it. If not, the job waits in the office, not on the floor, and purchasing knows exactly what to chase. The release check is also where lead time and purchasing meet. Material waits are usually the longest single gap in a job’s history, and they are the easiest to see in advance. A shortage list that nets confirmed demand against stock and open purchase orders, every morning, turns a surprise on the floor into a decision in the office. Our answer on [how to reduce material shortages](https://zudoworks.com/answers/how-to-reduce-material-shortages/) covers the habits behind it. ![ZMOS Procurement screen listing items short for confirmed sales orders and released jobs, with stock on hand, quantity on order and shortage for each item](https://zudoworks.com/Asset/img/manufacturing/zmos/zmos-procurement-shortages.webp) ZMOS screen The Procurement screen in ZMOS works out what is short for confirmed sales orders and released jobs, after stock on hand and open purchase orders, so material waits are seen before a job is released. Sample data. ### 3. Find the bottleneck and never let it starve Every factory has one step that limits output at any given time: often a press brake, a CNC machine, a welding bay, a paint line or a skilled assembler. Find it by looking for the step with the longest queue in front of it and the highest load. Then protect it: keep a small, deliberate buffer of ready jobs in front of it, schedule breaks and maintenance so it keeps running, move inspection before it so it never works on a bad part, and give it first call on material and people. An hour lost at the bottleneck is an hour lost for the whole factory. An hour saved anywhere else is often just more waiting at the bottleneck. ### 4. Understand why a busy machine makes everyone wait Most owners want every machine as busy as possible. Queueing theory explains why that lengthens lead time. Kingman’s formula, a standard result for a single machine, shows waiting time growing with utilisation divided by one minus utilisation. At 80% busy, a job waits on average about four times its own process time in front of that machine; at 95%, about nineteen times. The exact numbers depend on how variable your jobs and arrivals are, but the shape does not. ![Bar chart: waiting time as a multiple of the job’s own process time rises from 2.3 times at 70 percent utilisation to 19 times at 95 percent](https://zudoworks.com/Asset/img/manufacturing/reduce-manufacturing-lead-time-queue-curve.webp) Waiting time in front of one machine, as a multiple of the job’s own process time, from the standard queueing factor u ÷ (1 − u) with average variability. Real shops vary, but the shape is always the same: the last few per cent of utilisation cost the most waiting. The practical lesson: for the bottleneck, high utilisation is right, because it sets output. For every other step, some spare capacity is not waste. It is what lets jobs flow through without queueing. A shop that tries to run every machine at 95% will have long, unpredictable lead times however hard people work. ### 5. Cut setup times so batches can shrink Large batches are usually a response to long setups: if a changeover takes an hour, nobody wants to do it for ten parts. But large batches lengthen lead time, because every part waits for the whole batch at every step. Reducing setup time, by preparing tools and material while the machine is still running the previous job, standardising fixtures and writing down the setup steps, allows smaller batches without losing output. Start at the bottleneck, where an hour of setup saved is an hour of output gained. ### 6. Move work between steps in smaller transfers Even without changing setups, you can split a batch for moving. A batch of 200 parts does not have to finish laser cutting before bending starts. Moving the first 50 as soon as they are cut lets the next step overlap with the first. This one change often cuts days from long jobs. The same idea applies to paperwork: move finished jobs to the next step the same day, and mark the step complete when it happens, not at the end of the week. ![Before and after comparison: one batch of 200 moving as a block versus transfers of 50 so steps overlap](https://zudoworks.com/Asset/img/manufacturing/reduce-manufacturing-lead-time-batch.webp) Same order, same machines. Moving work in smaller transfer batches lets the next step start sooner, without changing the setup at all. ### 7. Plan purchasing from real demand Supplier lead time sets a floor under any job that needs bought-in material. You cannot change your supplier’s factory, but you can stop adding your own delays to it: raise purchase orders the day the need is known, not when the job reaches the floor; keep reorder points on the materials you use every week so they are simply there; and record honest due dates on open purchase orders so the release check is reliable. Our guide to [purchasing and reorder points](https://zudoworks.com/manufacturing/guides/purchasing-reorder-points/) explains how to set them. ### 8. Promise dates from load, not hope A short lead time is only valuable if customers can trust it. Quote delivery dates from the current load at the bottleneck and the supplier lead time for any material you do not hold, not from what the customer asked for. Record both the requested and the promised date on every order. Over time, the gap between them shows where you are losing work, and on-time delivery against the promised date becomes the measure of whether the plan is real. Our answer on [how to measure on-time delivery](https://zudoworks.com/answers/how-to-measure-on-time-delivery/) sets out the calculation. ## Seeing exactly where each job waits An average lead time tells you that jobs wait. Step-by-step time stamps tell you where. When each routing step is started and completed on the floor, by tablet or by scanning a barcode job card, every job builds its own history: released Monday 9:00, cut Monday 14:00, bending started Wednesday 10:00. The gaps between steps are the queues. Add them up by work centre over a few weeks and the bottleneck is no longer a matter of opinion. ![ZMOS work order showing routing steps laser cut done, bend running, assemble and pack pending, with planned hours and buttons to start and complete each step](https://zudoworks.com/Asset/img/manufacturing/zmos/zmos-job-routing.webp) ZMOS screen A work order in ZMOS. Each routing step is started and completed on the floor, which gives a time stamp for every step and shows exactly where a job is waiting. Sample data. This is also why work centres matter. If the whole factory is one block in your system, every wait looks the same. Set up each machine or team that has its own queue as a work centre, attach routing steps to them, and you can see waiting time and load per step. Our answer on [routings and work centres](https://zudoworks.com/answers/what-is-a-routing-and-work-centre/) explains the set-up in plain terms. ![ZMOS Work centres list showing each machine or team set up as a work centre](https://zudoworks.com/Asset/img/manufacturing/zmos/zmos-work-centres.webp) ZMOS screen Work centres in ZMOS. Setting up each machine or team as a work centre is what lets you see queues and hours per step rather than for the whole factory. Sample data. ## What to measure every week Pick a small set of measures, review them on a fixed day, and keep them for long enough to see a trend. Five are enough for most small factories: ![Table of five lead time measures with how to calculate each one and how often to review it](https://zudoworks.com/Asset/img/manufacturing/reduce-manufacturing-lead-time-measures.webp) Five numbers that tell you whether lead time is falling. All of them come from dates your system should already hold. Two cautions. First, look at the spread as well as the average: a factory where most jobs take five days and a few take thirty has a different problem from one where every job takes twelve. The jobs in the long tail are usually where the quickest lessons are. Second, do not let measures become targets for individuals. Lead time is a property of the whole system; blaming the person at the bottleneck for its queue will make the numbers worse, not better. Our guide to [manufacturing dashboards and KPIs](https://zudoworks.com/manufacturing/guides/manufacturing-dashboards-kpis/) covers how to show these numbers on one screen. ## Common mistakes that make lead times longer Most lead time problems in small factories come from well-meant habits. These are the ones to watch for: - **Releasing everything to keep people busy.** It looks productive and lengthens every queue. Busy people and short lead times come from a steady flow of ready jobs, not a pile of open ones. - **Expediting by shouting.** Pushing one job to the front delays every job behind it. A few genuine rush orders are fine; if more than a small share of jobs are “urgent”, the release rules have broken down. - **Buying a faster machine for the wrong step.** Extra capacity anywhere except the bottleneck adds no output and often adds more work in progress. - **Counting finished jobs at month end.** If steps are recorded late, you cannot see queues, and the measures arrive too late to act on. - **Treating supplier lead time as fixed.** It often hides your own delay in raising the purchase order, or a supplier who would deliver sooner with a better forecast. ## Where AI helps with lead time, and where it does not Lead time work is mostly about seeing the right numbers at the right moment, which makes it a good fit for an AI assistant that reads your own data. Useful questions include: which released jobs are waiting for material, which work centre has the longest queue this week, which jobs will miss their promised date at the current load, and what the average floor lead time was last month compared with the month before. Answering those in plain language each morning saves the time spent building reports by hand. Watch out The limits are the same as for any calculation. An assistant reading step times that were entered at the end of the week, or stock that is not posted when it moves, will give confident answers that are wrong. The safe pattern is the one we use in Zudo AI inside ZMOS: the assistant answers only from data the person is allowed to see, calculates and forecasts, and proposes changes, but every change is a proposal a person approves. Decisions about what to release and what to promise stay with people. ## Where ZMOS fits ZMOS, our manufacturing operating system for small factories, holds the dates and records a lead time programme depends on: work orders with routing steps and work centres, a Jobs board that shows what is released and in progress, steps started and completed on the floor with barcode job cards, a live stock ledger with reorder points, and a Procurement screen that nets confirmed sales orders and released jobs against stock and open purchase orders. Zudo AI answers questions such as which jobs are late or short from your workspace data, and proposes changes for a person to approve. ZMOS is not a finite capacity scheduler and does not optimise sequences automatically, and it has no dedicated lead time report today; you work lead time out from the dates on each job, or ask Zudo AI. A plant that needs minute-by-minute sequencing across many machines should pair a dedicated scheduling tool with its ERP. For most shops of 10 to 50 people, a WIP limit, a release check and protected bottleneck time deliver more than any optimiser. Startups and registered partners get six months free; see [ZMOS pricing](https://zudoworks.com/zmos/pricing/). ## An eight-week plan ![Eight-week plan to reduce lead time: measure, limit work in progress and check release, protect the bottleneck, then promise dates from load](https://zudoworks.com/Asset/img/manufacturing/reduce-manufacturing-lead-time-rollout.webp) An eight-week plan for a 10 to 50 person shop. Most shops see the first change in weeks 3 and 4, when the number of open jobs comes down. 1. **Weeks 1 and 2, measure:** work out order and floor lead time for every job finished in the last two to three months, count jobs open on the floor each morning, and note the five slowest jobs and why they waited. 2. **Weeks 3 and 4, limit and release:** set a cap on jobs open on the floor, a fifth to a quarter below today’s number, and introduce the release check so no job starts without its material. 3. **Weeks 5 and 6, protect the bottleneck:** find the step with the longest queue, keep a small buffer of ready jobs in front of it, move inspection ahead of it and start on its setup times. 4. **Weeks 7 and 8, promise from load:** quote new orders from the bottleneck load and supplier lead times, record requested and promised dates, and review the five measures every week. At the end of eight weeks, compare average floor lead time, jobs open on the floor and on-time delivery with your baseline. If lead time has not moved, the cap on open jobs is probably too loose or is being bypassed; tighten it before trying anything more expensive. ## Frequently asked questions There is no single good figure; it depends on the product, the routing and the market. A useful test is the share of lead time spent waiting. If hands-on work is a small fraction of the total, there is room to shorten lead time without adding machines. For each job, subtract the order date from the dispatch date for order lead time, and the release date from the last step completed for floor lead time. On average, lead time also equals jobs open divided by jobs finished per day, which is Little’s law. Not if the bottleneck always has work waiting. Output is set by the slowest step. Jobs open beyond what keeps that step busy only add queue time. Reduce work in progress gradually and watch the queue in front of the bottleneck. Lead time is the total time from order or release to finish, including waiting. Cycle time usually means the time to complete one unit or one step. Lead time is what the customer feels; cycle time is one of its parts. Because queues grow sharply as machines approach full utilisation. A machine at 95 percent busy makes jobs wait many times longer than one at 80 percent. Busy machines everywhere usually mean long, unpredictable lead times. No. Software shows where jobs wait, checks material before release and makes work in progress visible. The reduction comes from the rules you apply with it: a cap on open jobs, a release check and protecting the bottleneck. ## Sources Product capabilities change. We checked these pages in October 2026; confirm anything you plan around. 1. [Wikipedia: Little's law (Little, Operations Research, 1961)](https://en.wikipedia.org/wiki/Little%27s_law) 2. [Project Production Institute: Little's law, a practical approach](https://projectproduction.org/journal/littles-law-a-practical-approach-to-understanding-production-system-performance/) 3. [Wikipedia: Kingman's formula](https://en.wikipedia.org/wiki/Kingman%27s_formula) 4. [Zoho ERP help: manufacturing orders](https://www.zoho.com/en-in/erp/help/manufacturing-order/understanding-manufacturing-order.html) **Related services:** [BOM and production in ZMOS](https://zudoworks.com/zmos/production-and-bom/) · [Purchasing in ZMOS](https://zudoworks.com/zmos/purchasing/) · [Barcode job cards](https://zudoworks.com/zmos/barcode-job-cards/) ![Arunkumar V](https://zudoworks.com/Asset/img/team/arunkumar-v-avatar.webp) Written by [Arunkumar V](https://zudoworks.com/about/#leadership), Founder and CTO, Zudo Works Arunkumar builds Zoho Creator apps, Zoho CRM systems, Deluge automations and integrations for businesses in the US, UK, Australia, New Zealand and India. He received the “Master of Creator” award at the Zoho Creator Partner Hackathon 2025. Before founding Zudo Works in 2026 he worked at Teleperformance, at iOPEX Technologies and at a Zoho partner firm, and as a senior supervisor in manufacturing at Brakes India Private Limited, so he knows how a production floor runs. [Arunkumar V on LinkedIn](https://www.linkedin.com/in/arunkumar-v-cto)[About Arunkumar](https://zudoworks.com/about/#leadership) ## Related guides and blueprints Purchasing ### [Reorder Points, Safety Stock and Purchase Approvals: A Practical Setup](https://zudoworks.com/manufacturing/guides/purchasing-reorder-points/) How to set reorder points and safety stock for a small manufacturer, with formulas and worked examples, plus a purchase approval flow that keeps moving. Zudo Works development team · 6 min read Reporting ### [Manufacturing Dashboards and KPIs: What to Measure First](https://zudoworks.com/manufacturing/guides/manufacturing-dashboards-kpis/) Six manufacturing KPIs a small factory should track first, with formulas, a worked OEE example and rules for dashboards that people keep using. Zudo Works development team · 6 min read Inventory ### [Why Your Stock Records Are Wrong, and How to Fix Inventory Accuracy](https://zudoworks.com/manufacturing/guides/inventory-accuracy-for-manufacturers/) Why manufacturing stock records drift from reality, how to measure record accuracy, and the process and system changes that keep stock right. Zudo Works development team · 7 min read