A pre-aligned name and number set does not come back to you as editable text. It arrives cut, weeded and pre-aligned on a carrier, ready for one-step placement per color, which means the thing you approved has already been built. I keep that in view every time a team manager hands me a roster, because it changes what the document is: the spreadsheet you email is not a note to the design team, it is a build order that ends up as film sitting on a garment. A typo in it does not get retyped. It gets made.
Which is why I treat the roster, not the mockup, as the approval document. The mockup only shows that somebody read your list. Your list decides whether the mockup was right in the first place. Before any design file leaves your desk, every line has to carry four things: the jersey label that ties the row to one physical garment, the player name exactly as it will print, the number, and the size. Lock those four and the returned proof has something to be measured against. Leave out the label and you are signing off on a picture.
Map Each Jersey to One Name and Number
A bulk roster exists to manage names and numbers, and that is exactly the trap. A list of names in row order is not a mapping. As soon as somebody inserts a late registration in the middle, or two players share a surname, the row number stops identifying a garment. So the line has to be keyed on something that survives a copy-paste — the label — and the label can be anything the team already uses: the number written on a bag tag, a player ID, a bib assignment. What matters is that it is unique and that it does not move when the list is re-sorted.
Take two players named Ramirez, one at 12 and one at 21, both in medium. In row order those lines sit nine names apart, and to anyone reading quickly they are interchangeable. Add the label and they stop being interchangeable. Suppose I check a roster that has been through three rounds of copy-paste, one row duplicated and one deleted; the names can all look fine while the mapping is wrong, and the only thing that catches it is the label column failing to match the count of jerseys in the bag.
Spell Names Once in a Locked Column
Pick one column as the authority and let nothing else into it: capitals, apostrophes, diacritics, hyphenated surnames, the space in a two-part first name. Everything you want the supplier to ignore goes into a separate column marked clearly as not for print, whether that is a pronunciation note, a nickname the coach uses, or a size the player has outgrown. This sounds like paperwork for the sake of paperwork. It is not, because of where the text goes next: an uploaded roster comes back with each name individually aligned and ready for placement. A correction after that point is not a fix to a word, it is a rebuild of aligned artwork.
Numbers need the same single source. Decide once whether a number prints with a leading zero, whether a suffix belongs to it, and put that rule in the header rather than in a comment on row nine. Then the whole return on one locked column is a round taken off the schedule: spell-check after the first proof and you get two review cycles, lock before upload and you get one. That difference shows up as a delivery date, not as a quality argument.
Single Characters or Pre-Aligned Sets
This is the question that decides how the roster should be written, and it usually never gets asked. A pre-aligned set arrives already cut, weeded and aligned on a carrier, so each color is placed in a single step. In that format the roster is a build order: each line is bound to one garment and one placement run, and changing a line means that piece gets rebuilt. Loose single characters are a different animal. There the roster is a character list, the mapping happens on the floor at placement, and a wrong name can be swapped by reaching into the box for a different letter.
So the roster has to state which of the two you are buying, in the header, not in a side email. With a pre-aligned set, every line needs the label and the size, because those drive the build. With loose characters, what you need back from the supplier is a count per character per color plus a placement guide, and the per-jersey mapping moves to whatever document the decorator works from. Which of the two a given supplier offers is a question for that supplier to answer on your job; it is not a general claim about the category.
Walk the Roster Through Upload and Return
Here is the sequence as it actually runs. You finish the roster with labels, names, numbers, sizes and a format decision in the header. You hand it to the supplier's design system, either uploaded or typed into their roster tool. The alignment step sets each name individually against their template. What you get handed back is not a description of the job. It's the job, ready to run.
I check the alignment artifact first — not the spec sheet, not the pretty mockup — because the pieces inside it are what end up on the shirt. I want the names in the right order, the alignment matching the template, and the color layers split the way the placement runs split.
What does not come back is just as much a part of the answer. The bundle is decoration, not a commercial approval: it carries no price, no minimum, no capacity commitment and no lead time, and a tidy proof settles none of those.
Read the Returned Proof Line by Line
Read it against the labels, in label order, not in the proof's own order. Take the roster you sent, go down the label column, and for each label find the matching piece and check name against number against size. Reading in the proof's order is how shifted rows hide. Suppose you take the same proof and read it bottom-up as a second pass; a row that slipped by one becomes visible, because the name above no longer belongs to the number below. That pass costs minutes.
Two more items get checked in the same pass. Every name's alignment should match the template rather than the name next to it, and each placement run should carry its colors as separate layers. A name set in two colors is two passes, and a proof that merges them into one description will not tell you which pass comes first.
Imagine opening a returned set and finding one name sitting at the left edge of a block that is centered everywhere else. Nothing in the text is wrong; only the placement is, and it would have run exactly as shown. Sign off on a mismatch like that and the reprint is yours, not the supplier's, because the proof was built from the list you sent.
Freeze the Approved Proof Before Bulk Release
I version-tag the proof file I approve and keep the exact roster snapshot it was built from, in the same folder, with the date. Suppose a coach calls a few days later to say a player changed numbers. That change opens a new proof round against a new snapshot. It does not become an email that says same as before, just swap one number for another. Entries end up as aligned, pre-placed artwork, so a swap is a rebuild of that line, and a rebuild needs its own approval.
One boundary belongs in the checklist itself. A clean proof tells you the names and numbers are right; it says nothing about how the jersey performs in wear, whether a finish holds up, or whether any test report exists for the material, because none of that was tested in this loop. Those items, and any commercial term, are confirmed separately with each candidate supplier.
This checklist is not for the buyer who expects to swap names after the proof is signed, or who hands a rough list to a designer and hopes it comes back tidy. It is for the sourcing or QC lead who will lock a mapping jersey by jersey and read the returned proof against it, line by line, before releasing bulk.
Leave a comment