Lead time is a scheduling problem before it is a shipping problem

Ask a supplier how long something takes and you get a duration: thirty-five days for production, plus so many days on the water. It is a true answer to the wrong question. A buyer does not need to know how long the work takes. A buyer needs to know whether the goods will be available on the date the business planned around — and that is a different calculation, performed in the opposite direction.

Forwards and backwards

A forward schedule starts at the purchase order and adds durations until it arrives at whatever date it arrives at. A backward schedule starts at the date the goods must be on shelf or on site and subtracts, stage by stage, until it reaches the latest day an order can be placed. The two look similar written down and behave completely differently in practice.

A forward schedule produces an estimate, and estimates absorb slippage silently. A backward schedule produces a deadline for every party in the chain, including the buyer — which is why buyers are often reluctant to build one.

The commercial difference shows up most clearly in seasonal ranges. Landing stock in week one of a selling window and landing it in week six is the same lead time measured forwards and two entirely different outcomes measured backwards. The second one discounts.

A schedule with the missing stages restored

Written out for a typical Far East to United Kingdom movement, working backwards from the date the goods are needed, with the stages people leave out put back in.

• On-shelf or on-site date — the fixed point everything else is derived from.
• Retail or distribution centre processing, 1–2 weeks — goods-in, quality check, ticketing, allocation to stores.
• Inland transport and customs clearance, 3–7 days — assuming the declaration is right and nothing is selected for examination.
• Ocean transit, 30–40 days port to port — the number most people mean when they say lead time.
• Booking and port cut-offs, 7–10 days before the vessel sails — documentation, verified gross mass and container gate-in all close before the ship does.
• Production, 30–60 days — the supplier's own figure, and usually the only one quoted.
• Sampling and approval, 2–4 weeks — counter-samples, fit or shade approval, packaging artwork sign-off. Almost never in the quotation.
• Materials, 2–6 weeks where fabric, components or packaging carry their own lead time behind the production line, only partly in parallel with it.

Added together, a thirty-five day production quotation sits inside a programme of roughly nineteen to twenty-two weeks. The supplier was not misleading anybody. They answered the question they were asked.

Where the variance actually is

Buffer is usually applied as a flat percentage across the whole schedule, which is the least useful place to put it. Variance is not distributed evenly, and padding the stages that behave predictably only makes the programme longer without making it safer.

Production dates slip, but they slip within a range the factory knows and can be pressed on. Ocean transit is more reliable than its reputation for the crossing itself and much less reliable at the ends: a rolled booking on a full sailing costs a week or two with no warning and no recourse, and a container selected for examination costs days that nobody can predict.

The stage that most often destroys a schedule is sample approval — and it is the one entirely within the buyer's control. A counter-sample sitting three days on a desk has consumed three days of the critical path, and it will be recovered later by compressing production, which is the stage least able to absorb compression and the one where compression reappears as a quality problem.

The obstacles that are not variance at all

Part of what wrecks schedules is not uncertainty. It is a date in a calendar that everyone knows about and half the plans ignore.

Factories across much of East Asia close for one to three weeks around Lunar New Year, and the four weeks beforehand are the most contested capacity of the year. Ramadan affects output in parts of South and Southeast Asia. European August thins out haulage, and the fortnight around Christmas thins out clearance. None of this is a surprise to anyone. All of it belongs in the schedule as a fixed obstacle rather than as the subject of an apology in February.

What to ask, and what to write down

1. 01Ask for an ex-factory date rather than a lead time. A date is a commitment; a duration is an estimate that belongs to nobody.
2. 02Ask what that date assumes — materials in stock, artwork approved by when, deposit received by when — and write the assumptions next to it.
3. 03Fix your own approval gates as dates in the same schedule, with your name against them.
4. 04Book freight against the ex-factory date with the cut-offs built in, not against the day the goods are ready.
5. 05Agree in advance what happens if the date moves: partial shipment, air freight for the balance, or a revised on-shelf date. That is a commercial conversation while there is still time, and an argument once there is not.

So the first question is not how long something takes. It is what date the goods are needed, and then whether anybody in the chain has committed to a schedule that produces it.