One brief, three correct products

The situation

Ask three factories for a waterproof jacket and you will get three different garments. All three will be honest and all three will be correct. One seals every seam. Another seals only the seams water runs down. The gap between their prices then looks like a bargain, and mostly it is not. This lesson writes the kind of description that makes quotations comparable. It also counts everything else that lands on a shop developing a product for the first time.

4 February. The Fellgate jacket has been approved in principle and the buyer sends a brief to three factories. It is one page, and it is the brief a competent buyer writes. Every line on it is true, none of it is careless, and it reads like this.

• Men's three-season waterproof walking jacket, hooded, hip length
• Waterproof, taped seams, breathable
• Fully lined
• Two-way front zip, water-resistant
• Packs into its own pocket
• Target landed cost GBP 36.00, 3,600 units, three colourways

On 22 February three quotes come back: GBP 23.10, GBP 30.20 and GBP 38.40 free on board. FOB, or free on board, means the price of the goods loaded onto the ship, with the buyer paying the freight from there. The buyer's first reaction is that Rach Gia Apparel is a third cheaper than Batangas Technical Apparel, and that this is a very good day.

It is not. The three factories have quoted three different garments, and every one of them satisfies the brief.

The same six lines, read three ways

Read the "taped seams" row again. The brief said taped seams. Rach Gia taped the seven seams above the waist that water actually runs down. That is a real construction, real jackets ship with it, and it is exactly what "taped seams" means on a great many garments. Ninh Giang taped all nineteen. Neither factory has misread anything. The brief did not say how many.

Now the row that matters most. Rach Gia quoted no test method at all, so the 3,000 mm figure could have come from anywhere. Both of the others named ISO 811, the hydrostatic head test: it pushes water against the fabric under rising pressure and records the height of water the cloth holds back before it lets a drop through. That single word changes the quote, and why it does is the point of this whole lesson.

A performance claim has three parts, and a brief usually has one

A claim is a property, a method, and a value. "Waterproof" is a property. On its own it is a wish.

• The property alone — waterproof — lets a factory choose anything from a shower-resistant coating to a three-layer laminate.
• The property and a value — waterproof, 10,000 mm — is better and still not enough. Two laboratories measuring "waterproofness" different ways will hand you different numbers off the same cloth.
• The property, the method and the value — not less than 10,000 mm to ISO 811 — is a specification. It is the first of the three that a factory can price and a laboratory can settle.

And the value has to be attached to the right thing. A hydrostatic head measured on the fabric says nothing about the seam, and a jacket leaks at its seams long before it leaks through its cloth. So Ludgrove's closed specification carried two numbers on two objects: not less than 10,000 mm on the shell, and not less than 8,000 mm on a taped seam after five domestic washes. The test methods themselves — how they work, why two of them disagree, what a laboratory can and cannot settle — belong to course 6.3, which owns them. What belongs to you is knowing that the method is half the specification, and that a brief without one is not a brief.

The same discipline applies to "breathable", which is measured several ways that are not interchangeable, and to "water-resistant zip", which means a zip with a coated tape and no more.

The spread was the specification, not the suppliers

Ludgrove rewrote the brief. It named the methods and the values, said nineteen seams, named the lining by construction and weight, and re-issued it on 8 March. The re-quotes came back on 18 March.

The apparent spread of GBP 15.30 across 3,600 units looked like GBP 55,080.00 of buying opportunity. GBP 43,560.00 of it did not exist. It was the difference between three garments. Once all three factories were quoting the same garment, the real spread between suppliers was GBP 3.20.

Two things follow, and the second one is the useful one.

Rach Gia was never a cheap factory. It was a cheap garment. The GBP 6.30 it added on the re-quote is exactly the specification that Ludgrove's own brief had failed to ask for, and that Ludgrove's own claim depended on. Had the buyer placed the order on 22 February at GBP 23.10, nobody would have been deceived, and Ludgrove would have sold a jacket that did not do what its label said.

And Batangas came down GBP 5.80. A closed specification does not only reveal who was quoting cheap. It also stops a good factory pricing in the risk of a brief it cannot bound. When a supplier does not know what you will finally demand, the unknown goes into the price.

The work the brand used to do, itemised

Braemount's GBP 62.00 included a great deal of work that nobody at Ludgrove had ever seen. Here it is, as cash, for one style in one season.

Grading, in the first line, means working the pattern up and down into the other sizes. A lab dip is a small trial dyeing sent for colour approval.

Two lines in that table are the ones buyers forget.

The 150 hours. Forty-one of the buyer, seventy-six of the garment technologist, thirty-three of an assistant. It is the second largest line on the page, and it does not appear on any purchase order, which is exactly why it goes missing. It is also the line that does not scale. A second own-brand style does not need a second trade mark, but it needs another 150 hours.

The trade mark. Fellgate is a name Ludgrove has to own before it puts it on a garment, and a registration is granted for named classes of goods. A mark registered for clothing does not cover the rucksack the same range wants next year.

Sample rounds, who pays, and the days that are yours

The charges come to GBP 1,235.00, which is trivial. The calendar is not. Eighty-eight days of the development sat in the sample loop, and 22 of them — a quarter — were Ludgrove's own review time, not the factory's making time. A retailer new to own-brand reads its own days as free, because no invoice arrives for them. They are the days that later have to be recovered from the shipping window. What a day of a buyer's silence costs the person waiting on it is course 27.4's subject, from the giving side. The point here is only that those days are now on your calendar, and they were not before.

What the retailer now has to write down

This is the sentence to take away from the lesson. Own-brand does not move a product from one supplier to another. It moves a body of work from the brand to you, and most of that work is documents.

• A base size and a size chart, and grade rules between the sizes. Braemount's fit was Braemount's asset. Yours does not exist until you write it.
• A garment measurement specification with tolerances — the point at which a jacket is off-spec rather than merely different. Course 4.6 covers reading a tech pack. You are now the one writing it.
• Which sizes you approve on. Approving the base size and nothing else is how a range fits beautifully at medium and badly at both ends.
• A quality standard and an inspection level for the goods you receive. Track 6 owns the AQL machinery, which is the sampling system that decides how many pieces you check and how many faults you accept. Owning the machinery is not the same as choosing the numbers, and the numbers are now yours.
• A restricted substances list — the chemicals that must not be in the product — and a testing schedule. Lesson 3 takes that up.
• Care instructions that are a promise about the whole assembly, not a copy of the shell mill's card.
• A trim and component approval list — every zip, cord, tape and label named by supplier and reference, so that a substitution is visible.

The calendar, and the thing it makes impossible

Ludgrove placed PO LD-7742 on 24 March. The jackets reached the distribution centre on 22 September: 26 weeks, exactly. The season opened on 13 October and closed at the end of February, twenty weeks.

Set those two side by side and the structural difference is finished in one line. A repeat ordered on launch day would arrive six weeks after the season ended. The branded route's repeat took five weeks and was still an option in January. There is no version of the own-brand route where a repeat is possible, and no amount of goodwill from the factory changes it, because the constraint is the fabric, not the sewing. How these dates are set, anchored and defended is course 16.3's subject, and course 7.1 covers the same calendar from inside the factory.

The change on 6 May that nobody flagged

The round-four sample arrives on 6 May in bulk fabric. The buyer handles it, compares it with the Braemount jacket on her desk, and writes one comment.

Shell feels boardy next to the Ridgeway. Can we soften it?

That is a good comment. Hand — how the cloth feels between your fingers — is the first thing a customer judges and the last thing a factory volunteers. Ninh Giang asks the laminator, who proposes using a thinner face yarn and changing the finish. Cost effect: minus GBP 0.14 a jacket. The revised sample arrives, the hand is right, and the technologist approves it on 13 May on colour, hand and construction — correctly, on all three.

Nobody did anything wrong. Lesson 3 is what that comment cost.

Act as a garment technologist who writes retailer specifications for a living. I am going to paste a product brief, and I want it turned into something three factories cannot legitimately read three different ways. Here is the brief: [PASTE IT EXACTLY AS SENT]. Product category [CATEGORY], intended retail [AMOUNT], target landed cost [AMOUNT], quantity [NUMBER], season [SEASON]. Now do the following. First, list every ADJECTIVE and every performance word in the brief. For each one, tell me the property it names, the test method that would settle it, and whether the brief gives a value. Mark the ones where the brief gives a property with no method and no value. Those are the lines that will produce different quotes. Second, for each performance claim, tell me WHAT the value should be attached to: the fabric, the seam, the finished garment, or the garment after a stated number of washes. Then say where a value attached to the wrong thing would mislead me. Third, list the construction decisions the brief leaves open, of the how-many and which-ones kind. Give the cheapest and the dearest honest reading of each, and roughly what the difference is worth per unit. Fourth, write me the closed version of the brief. Fifth, list what the closed brief still leaves to the factory on purpose, because a specification that closes everything is a specification nobody can quote. Sixth, if I already have quotes, tell me which differences between them are supplier differences and which are specification differences, and warn me where you cannot tell from what I have given you. Do not invent a standard reference you are not sure of. Where you are unsure, describe the method in words and tell me to have it confirmed.

What you do on Monday

Take a brief you have already sent and mark every adjective in it: waterproof, breathable, soft, durable, premium, heavyweight, colourfast. For each one write the property, the method and the value, or delete the word. If you cannot supply a method, that is the line to ask a technologist about before you ask a factory about it.

Then find out how many days of your last development sat on your own desk. Not the factory's. Yours. It is a number nobody at a retailer has ever measured, and it is usually a quarter of the calendar.