How to Measure and Reduce Supplier Lead Times
Split a lead time into its four parts, fix the one with the widest spread, and check that the cut held.
You reduce supplier lead times by splitting them into four parts, order processing, manufacturing, transit and receiving, and fixing the part with the widest spread in days, not the longest average. A 3-day receiving step that swings between 1 and 9 days can cost you more safety stock than a steady 22-day factory lead time. This guide shows how to baseline each part, rank them, cut the one that drives your expedites and stockouts, and check that the cut held. How to measure a total lead time from your receipts is covered in supplier lead time.
What are the components of a supplier lead time?
A supplier lead time is four clocks running in sequence. Each one starts and stops on a timestamp you either already have or can get.
| Component | Window | Where the timestamps live |
|---|---|---|
| Order processing | PO sent to supplier confirmation | PO issue date and acknowledgement date |
| Manufacturing | Confirmation to shipment | Confirmed date and ship notice or ASN |
| Transit | Shipment to your dock | Ship date and goods receipt |
| Receiving | Dock to stock | Goods receipt and stock posting |
Pull 10 to 20 recent POs per supplier and work out the average and the standard deviation of each part. Most ERPs at smaller manufacturers record only the PO date and the receipt date, so manufacturing and transit blur together. A ship confirmation from the supplier, even by email, splits them.
Why does lead-time spread matter more than lead-time length?
A long but predictable lead time is a planning problem: you order earlier. An erratic one is a stockout problem, because safety stock grows with the spread of lead time, not with its length. The four parts add up to the spread of the whole lead time like this, as long as they vary independently:
That total spread, σLT, goes straight into safety stock. Hover over or tap any symbol to see what it stands for.
Take two suppliers of the same part, used at 100 units a day with a standard deviation of 30, held at a 95% service level. Supplier A quotes 10 days and delivers anywhere from 4 to 20. Supplier B quotes 18 days and lands within a day of it every time.
| Supplier | Delivers in | Average | Standard deviation | Safety stock |
|---|---|---|---|---|
| A | 4 to 20 days | 12 days | 4.6 days | 779 units |
| B | 17 to 19 days | 18 days | 0.6 days | 230 units |
The faster supplier needs about 3.4 times the buffer, and its real average is 12 days, not the 10 it quotes. B is easier to plan: you order earlier, hold less stock and rarely expedite. The safety stock calculator runs these numbers for your own parts.
Which lead-time component should you reduce first?
Rank the four parts by their standard deviation in days, and fix the one at the top. Each part's share of the total spread is its standard deviation squared, divided by the sum of all four squares, so the widest part frees the most buffer when you tighten it. The longest part is often the most predictable: an ocean leg that lands within a day every time is easier to plan around than a dock that sometimes takes a week.
Fixing the widest part usually pays back in weeks, because it is often inside your own walls or in the supplier's paperwork, not their factory. Cutting the longest part usually means renegotiating with the supplier and waiting a quarter for new terms to show up in receipts.
Say a part's lead time averages 32 days, and your last 20 receipts break down like this:
| Component | Average | Standard deviation | Share of spread |
|---|---|---|---|
| Order processing | 2 days | 1.0 day | 9% |
| Manufacturing | 22 days | 2.0 days | 34% |
| Transit | 5 days | 1.0 day | 9% |
| Receiving | 3 days | 2.4 days | 49% |
Receiving is the shortest part with the widest spread, because the dock is short-staffed on Fridays and receipts sit over the weekend. Fix the dock schedule so receipts post in a day with a spread of half a day, and σLT drops from 3.4 to 2.5 days. At the usage and service level above, safety stock falls from 629 to 492 units, 22% less, without a call to the supplier. Then rank again.
Length wins when the longest part also has a wide spread, or when the whole lead time runs past how far ahead you can see real demand, so you order against a forecast. A shorter lead time then cuts the forecast error too.
How do confirmed dates change your effective lead time?
Track three dates per PO line: the date your ERP planned, the date the supplier confirmed, and the date the part arrived. Until a supplier confirms, your ERP plans against the lead time in master data, and you commit capacity to a date nobody agreed to. The gap between confirmed and actual is the spread you plan against, and it is usually wider than the gap between planned and confirmed. Purchase order acknowledgement covers turning a PO into a confirmed date.
A confirmation from eight weeks ago is a stale date, not a commitment. On long-lead parts, ask the supplier to reconfirm on a fixed schedule, weekly or every two weeks, so a slipping ship date shows up weeks before it turns into an expedite. When dates slip again and again, take it to a corrective action plan, and when the supplier goes quiet, follow the steps for a supplier that isn't responding. Late confirmations are often an early sign of supplier capacity risk.
When should you buffer instead of chasing a shorter lead time?
Run the numbers both ways. A buffer costs carrying cost, obsolescence risk and working capital every month, on every unit. A reduction project costs buyer hours, supplier negotiation and process changes, mostly once. Compare a year of carrying cost on the extra stock with the hours the project will take.
Buffer when:
- The part is cheap, high-volume and stable in demand, with little risk of a redesign.
- Expedites and premium freight on stockouts already cost more than carrying the stock would.
- The part is custom or single-source, and you have no leverage to change how the supplier works.
Reduce when the part is expensive, or when the spread comes from a supplier or a step you deal with on every order. Don't buffer parts with lumpy, erratic demand: the formula assumes steady use, and you end up with excess stock rather than protection.
How do you check that a lead-time reduction held?
A supplier promising a shorter lead time is not the same as delivering one. Compare a 90-day window before the change with a 90-day window after it, part by part, using medians rather than averages, because one expedite drags a mean. If a supplier ships six orders a month, 90 days gives you about 18 receipts: enough to see the median move, not enough for tight limits.
Then plot each part on a control chart, with limits from the before window, and flag any receipt outside them. Points drifting toward a limit count as a slip even when the average still looks fine. Check monthly; a gain that holds for one quarter only is noise.
How do you run this without adding headcount?
The judgment calls stay with your team: which part to work on, whether buffers still match reality, and the conversation with the supplier. A monthly hour reviewing the breakdown is enough for most suppliers.
The confirmations and status checks that feed the breakdown are where the hours go. When a PO has no confirmed date, or a delivery is due and nothing has arrived, a readiness signal opens a task, and Cartograph hands it to your team or Scout Agent, which emails and calls the supplier within the permissions your team sets. The platform is free for unlimited team members; Scout Agent is priced per successful outcome. Cartograph will not negotiate with your suppliers or fix a wrong lead time in your master data; someone on your team still owns both.
Questions teams ask
What is a good supplier lead time for a small manufacturer?
There is no universal number. A good lead time is shorter than the horizon over which you can see real demand, and steady enough to plan against. A steady 30-day lead time beats an erratic 15-day one, because the spread sets how much safety stock you carry. Start from your own baseline.
How do you calculate lead-time variability?
Pull 10 to 20 recent POs per supplier, split each into order processing, manufacturing, transit and receiving using ERP timestamps, and take the standard deviation of each part in days. The total spread is the square root of the sum of their squares.
Which lead-time component should you reduce first?
The one with the largest standard deviation in days, not the longest average. A receiving step that swings between 1 and 9 days can force more buffer than a steady 22-day factory lead time. Fix the widest one, rank again, and repeat.