Who Signs What Before 500 Polos Get Made
Most company apparel problems were approved in writing by the buyer. A vendor builds what was signed off, which makes the approval the real work, and four completely different objects get put in front of a buyer under one word: sample. The four are the digital proof, the blank garment, the sew out or strike off, and the pre production sample. Only the last one is the product. The only signature that matters is the one on the first finished shirt in the real colorway with the real mark on it, and the three objects before it exist so that shirt is right the first time.
Four Different Objects, One Word
Four objects travel under the word sample, and each answers a question the others cannot. A digital proof is a picture. A blank is a garment with nothing on it. A sew out or strike off is a mark with no garment around it. A pre production sample is the shirt itself. Hand a buyer any one of the first three, call it approval, and the buyer has signed for something that is not what ships.
The digital proof
ProvesPlacement, spelling, whether the artwork is complete, the size breakdown on the order.
Does not proveThread color, stitch density, how a mark behaves on a stretch knit.
The blank
ProvesFit, the size run, the fabric in the hand, the real color of the ground.
Does not proveAnything at all about the logo, because there is no logo on it.
The sew out or strike off
ProvesThread color against the ground, density, whether fine detail and small type survive, whether the mark puckers.
Does not proveFit or placement on a body, because there is no body under it.
The pre production sample
ProvesAll of it at once. Real colorway, real mark, real placement, real size.
Does not proveThe bulk. It is one garment, under one light, off one roll.
The order matters as much as the list. Proof first, because spelling and placement are cheapest to fix on a screen. Blank second, because fit and the real ground color decide the size run before a single stitch goes in. Sew out third, because thread behaves nothing like pixels and you want to learn that on a swatch. First finished shirt last, because it is the only object that carries every decision at the same time.
Most programs stop at one or two. The proof gets signed because the vendor needs a signature to start. The blank gets ordered because it is free or close to it. The sew out gets skipped because nobody asked for it, and the pre production sample gets skipped because it costs days. Those two skips produce the box nobody wants to open.
A Picture on a Screen, No Thread In It
A digital proof proves placement, spelling and artwork completeness, and it proves nothing about thread. It is a flat image of your mark laid over a flat image of a garment. It will catch a misspelled tagline, a crest sitting too low, a logo supplied at the wrong proportions, a size breakdown that adds up to the wrong headcount. Those are real catches and they are worth the round trip.
What a proof cannot do is tell you what color the thread will be against the fabric you chose. Underground Shirts states the limit in its own proof guidance: "Most clients receive a digital proof, which may not fully capture colors or exact print placement." The same guidance adds that "Colors can vary between screens and fabrics." That is one company describing its own process, not a PILLAR number, and it names the physics every decorator works inside. Your monitor emits light. Thread and fabric reflect it, unevenly, at whatever angle the shoulder happens to be turned.
A screen has no thread on it, and no knit under the thread.
So approve the proof for what it is good at and then stop. Read every word, including the words inside the logo lockup, out loud. Check placement against a measurement rather than a feeling, a stated distance down from the shoulder seam and in from the placket, so the same number can be used on next year's reorder. Confirm that every garment, colorway and size on the proof matches the sheet you built. Then put the pen down, because the next question is a thread question, and it belongs to the method that will actually apply the mark.
The Blank First, Before Any Logo
A blank is the fastest way to be right about fit, and it is the sample that tells you nothing about your logo. An undecorated garment in the real colorway puts the fabric in your hand, shows the ground color under the light your people actually stand in, and lets you run the size question properly, which means one person per body type wearing the garment rather than four people reading a chart.
The industry is honest about this one almost by accident. CustomInk tells buyers in its help center to expect "an unprinted sample at your door within a week", and in the same breath that "printed samples do have a non-refundable $100.00 fee associated". Those are CustomInk policies and not PILLAR numbers, and the shape of them is the useful part. The convenient sample the market hands out by default is the blank, the second object on the list, and the decorated one is a separate thing you have to ask for and pay for. The whole industry is quietly telling buyers that the easy sample proves the garment and not the mark.
Use the blank for the decisions only a body can make. Does the collar still stand up at the end of a day. Does the hem sit right untucked. Does the person who hates company shirts stop hating this one. Write the size run off those bodies, then keep the blank in a drawer, because when the first finished shirt arrives you will want both in the same hand under the same light.
The David Polo is the blank most corporate programs here start from, at $76 in Navy, Black and White. It is 84 percent nylon and 16 percent spandex, four way stretch, sweat wicking, cut to a classic athletic fit with a firm collar, a U cut hem and a pre shrunk body. Order two colorways rather than one. A ground color you were sure about on a screen often loses the argument the moment it is next to its neighbor on a table.
Thread on the Real Knit, Not on Paper
A sew out is the mark alone, stitched on the fabric you actually ordered, and it is the only object that can answer a thread question. It proves thread color against the ground, stitch density, whether fine detail and small type survive at the size you asked for, and whether the mark puckers or distorts on a four way stretch knit instead of sitting flat the way it would on a woven. It proves nothing about fit or placement on a body. For an applied graphic the same object is called a strike off and it does the same job for reflective, tonal and metallic transfer.
The stretch is where programs get surprised. A logo drawn for a business card and a logo that has to hold together small on a knit that moves are two different files. Thin strokes close up. Small counters fill in. A tight registration between two thread colors drifts by a hair and the eye catches it immediately. None of that is a vendor failing, it is what thread does on fabric, and the sew out exists so you find out on a swatch instead of on 240 shirts. If a specific brand color is the whole point of the program, settle how that color gets matched on a polo before the sew out is run, not after.
White, $76
Ask for your sew out on white if you can only have one. A white ground is where a thread color reads truest, because nothing underneath is pulling the thread warmer or cooler or darker. The same spool on black reads a full step deeper than it is, and a navy that was approved on navy can disappear into its own garment.
That makes white the honest test and the hard test at once. If the mark holds its color and its detail on white, you know what the thread actually is. Then run the dark grounds against it and decide whether the mark needs a different thread on those, which is a decision, not a problem.
One Finished Shirt, in the Real Colorway
The pre production sample is the first finished shirt, real colorway, real mark, real placement, real size, and it is the only one of the four objects that is the product. Everything before it is an instrument. This is the thing itself, and it is the only object worth a signature.
Judge it against the other three rather than on its own. Lay the proof beside it and check placement against the measurement you approved. Lay the blank beside it and check that the ground color did not shift when the garment went through decoration. Lay the sew out beside it and check that the mark on a finished body panel reads the same as the mark on a swatch. Then put it on a person, because a crest that looked correct flat on a table can sit wrong on a chest, and a sleeve hit that was fine in a proof can land on a seam.
This is the step the programs with the most at stake never skip. The apparel PILLAR builds for AO and the WM Phoenix Open goes through a finished first piece before anything is cut in volume, and the reason is not caution. It is that a correction on one shirt costs an afternoon and a correction on a full run costs the date.
| What you are sent | What it proves | What it cannot prove | Sign it? |
|---|---|---|---|
| Digital proof | Placement, spelling, artwork completeness, the size and color breakdown on the order. | Thread color, density, how the mark sits on a knit that stretches. | Yes, and understand you have approved spelling and placement only. |
| Blank sample | Fit, the size run, fabric in the hand, the true ground color. | The logo, in every respect, since there is none on the garment. | Yes, for fit and colorway, after real bodies have worn it. |
| Sew out or strike off | Thread against the ground, density, survival of fine detail and small type, puckering. | Fit, placement on a body, how the finished garment hangs. | Yes, and know it locks the stitch file for this order and later ones. |
| First finished shirt | The product. Colorway, mark, placement and size together on one garment. | The full run, which comes off other rolls and other lots. | Yes. This is the signature that matters. |
| Printed mockup on a model | How the program will photograph, useful for internal buy in. | Everything physical. It is a rendering with a person behind it. | No. Approve the proof instead and keep this for the announcement. |
| Reorder off a locked stitch file | That the mark will repeat exactly as approved, stitch for stitch. | That the new fabric lot reads identically to the first one. | No new signature on the mark. Ask about the lot instead. |
Three Signatures Inside Your Own Company
Collect three signatures inside your own company before you send one approval out, because the person who signs the vendor's form is usually neither the person who owns the mark nor the person who has to wear it. No vendor is going to tell you this, since the vendor only needs one name on the line.
The first signature belongs to whoever owns the brand mark, marketing or the owner. They decide thread color, mark size and whether the secondary lockup is allowed on a sleeve. The second belongs to whoever will actually wear the garment on a floor, and that is one person per body type, holding the blank, in the building, not a reply to a group email. The third belongs to whoever pays, because a change requested after approval arrives as an invoice and nobody enjoys that conversation twice.
Running those three in parallel inside one week beats running them one at a time across three. Send the blank around with a single question attached to each person rather than a general request for thoughts. General requests for thoughts produce opinions about the color of the logo from the person who was supposed to be checking the hem. If you are standing up a program from nothing, it is worth knowing what to pin down with a vendor before any of these three see anything at all.
Two Clocks, Both Starting at Your Approval
There are two clocks and they are nowhere near each other. Logo embroidery on stock polos runs 7 to 10 business days from proof approval. A full custom program built from the fabric up runs about 45 days door to door, with production alone taking 30 to 35 of those days. Picking the wrong clock for your program is how a date gets missed before anyone has done anything wrong.
Both clocks start at your approval, the proof for embroidery and the artwork for a build, not at the purchase order. That is the single date on the whole project that you control. Every approval round you ask for is days, and those days come out of the front of the clock rather than the back, because the back end is machine time and freight. Ask for the samples you need, ask for them in one batch, and understand that the batch has a price in days. If the date is already tight, look at what a compressed build actually gives up before you decide which approvals to drop.
One Light, One Dye Lot, One Roll
An approved sample is approved under one light, on one dye lot, from one roll. It is the best instrument a buyer has and it is not a guarantee of the bulk. PILLAR does not publish a tolerance figure, a color difference number, a colorfastness rating or a wash count, so you will not find one here, and you should be careful with any vendor who hands you one without saying how it was measured.
What can be committed to in writing is process. If a color has to hold across a multi wave order, ask for the bulk to be cut from a single dye lot and get that in the order confirmation rather than in a phone call. Ask for the approved placement measurement to be written on the order. Ask for the approved stitch file to be held and reused by file name on reorders. Ask that the approved sample itself be retained so a later question has a physical reference rather than a memory.
Look at the sample in the light your people live in too. A navy approved under warm office light and worn under a parking lot at six in the evening is two different navies, and that is not a defect in anybody's shirt.
Your Signature, and What It Locks
Each signature locks a different amount of the project, and the amount grows fast. Signing a proof locks placement and spelling. Signing a sew out locks the stitch file, which means it locks thread color and stitch density for this order and for every reorder run off that file. Signing the first finished shirt locks the whole product, and after that a change is a new order with a new clock, not a revision.
That escalation is also why each new garment in a program is its own approval. PILLAR builds custom fabric woven or knit to spec and dyed to an exact color, all over print from original repeat patterns, embroidery including chest crests, sleeve hits, chenille and 3D puff, applied graphics in reflective, tonal and metallic transfer, private label neck tags, trim work, headwear, and packaging with branded insert cards and name level kitting. Every one of those carries its own approval, and none of them inherits a signature from the polo.
Blue, $98
A layer is a separate approval, not a line added to the polo order. A logo file set up for a flat knit has to be re set for a lofted face, so a program that adds a quarter zip signs a second sew out. The mark can also need a different size and a different position once there is a zipper and a placket in the way.
Sizes do not carry over either. You cannot derive the layer size run from the polo size run, because people wear a layer over something. Put the layer on the same bodies that wore the blank and let them tell you, then sign the two programs separately.
One more habit worth keeping. Write down what each signature covered on the day you gave it, in one line, in the order confirmation. A program that can say which object was approved and when never has the argument about who changed what, and the whole polo range stays reorderable for years off that record.
The Questions Buyers Ask Before Signing
These are the questions that come in from buyers in the week before an approval is due, with the short answers.
Is a free sample the same as a sample with my logo on it?
No. The fast, low cost sample most vendors offer is an undecorated blank, which proves fit, fabric and ground color. A decorated sample is a separate request and usually carries its own cost and its own days.
Can I approve a digital proof and skip the rest?
You can, and you will have approved a picture. A proof settles placement, spelling and the order breakdown. It cannot settle thread color, density or how the mark behaves on a knit that stretches.
What does signing a sew out actually lock?
The stitch file. That means thread color, stitch density and the mark itself are locked for this order and for any reorder run off the same file, which is why a sew out is worth a careful look rather than a quick yes.
How many days does sampling add to my timeline?
Both clocks start at your approval, so every approval round is days taken off the front of the schedule. Logo embroidery on stock polos is 7 to 10 business days from proof approval, and a full custom build is about 45 days, with 30 to 35 of those in production.
Will my bulk order match my approved sample exactly?
An approved sample is one garment, under one light, from one dye lot and one roll, so treat it as the best reference available rather than a guarantee. If a color has to hold across a multi wave order, ask in writing for the bulk to be cut from a single dye lot.
Do I need a separate sample for a quarter zip or a jacket?
Yes. A layer is its own approval, because a file set for a flat knit has to be re set for a lofted face and the mark may need a new size and position. The layer size run also cannot be derived from the polo size run.
Tell PILLAR the event, the headcount and the date. We will tell you which approvals that date can carry, which samples are worth the days, and which of the four objects you actually need in your hands before you sign anything.
