How to Choose a Bag Logo Method: Fabric First, Quantity Second, Looks Last
You picked embroidery off a mood board. It looked expensive, in the good way.
Round three of sampling is when someone finally says it out loud: your face fabric is TPU-coated, the logo is going on the front panel, and stitching it there means roughly eight thousand needle holes through your waterproof layer. Right where the rain hits.
That conversation happens constantly, and it happens late, because almost everyone chooses a logo method in the wrong order. They start with what looks right and then go hunting for a fabric that can survive it.
Run it the other way.
Your fabric eliminates about half your options before you’ve made a single decision. Your quantity eliminates half of what’s left. How the logo will fail eliminates the next one. Aesthetics gets whatever survives — and that is the correct order, because the first three constraints are physics and the fourth is taste.
This is the sequence, with the numbers.
Why does your fabric choose the method, not you?
Because a logo is a bond between two materials, and most fabric-and-method pairs don’t bond. Coated fabrics can’t be punctured without losing what the coating was for. Slick PU won’t hold screen ink. Nylon melts before a foil die gets hot enough. None of that is negotiable, and all of it is decided by a line in your tech pack that was written for a completely different reason.
| Fabric | Works | Fights you | Why |
|---|---|---|---|
| Cotton canvas / cotton twill | Screen print, embroidery, patches | Deboss (no depth to hold) | The friendliest substrate in bags. Ink sits, thread bites. |
| Nylon / polyester (uncoated) | Screen print (needs adhesion-grade ink), embroidery, heat transfer | Standard textile inks — they scuff off | Synthetics are slick. You need a nylon-specific ink system, and you need to say so. |
| Coated / waterproof (TPU, PU-coated, laminated) | Heat transfer, printed patch, welded/RF logo | Embroidery (punctures the coating), high-temp foil (melts it) | Every stitch is a hole. See below. |
| PU / faux leather | Deboss, emboss, hot stamp, metal plate | Screen print (poor adhesion on slick finishes) | Pressure-based methods, not ink-based. |
| Genuine leather | Deboss, emboss, hot stamp, metal plate | Embroidery on thick hides; screen print | Same as PU, plus leather is expensive to test on. |
| RPET / recycled poly | Screen print, heat transfer, woven label | High-heat methods on thin recycled weaves | Behaves like polyester but is often thinner. Test. |
The needle-hole problem nobody in bags writes about
Embroidery does not sit on fabric. It goes through it. A logo of any size is thousands of punctures in a concentrated area, and on a waterproof or water-repellent fabric that is exactly what it sounds like. The apparel industry has known this forever — SanMar’s own decorator guidance says it plainly, and there’s a whole product category (heat-activated laminate films like Thermoseal) that exists to seal the back of embroidery on rain gear.
It’s solvable. Laminate backing, heat-pressed over the reverse of the stitching, closes the holes.
But it has to be specified, it costs money, it adds a step, and if nobody raises it until the third sample, it isn’t in your quote. Worse, on a bag the geometry is against you: on a jacket the embroidery is on a left chest, small and rarely rained on directly. On a backpack it’s front and centre, on the panel most exposed to weather, above a compartment holding a laptop.
If the words “waterproof” or “water-resistant” appear anywhere in your product copy, and the word “embroidered” appears anywhere in your tech pack, someone needs to have this conversation before sampling. Not after. Our guide to how bag fabrics actually behave is the other half of this decision.
Why is “cost per unit” the wrong number?
Because every logo method has a one-time tooling charge, and at low volume the tooling is the cost. The method with the lowest per-unit price and the method with the lowest total cost at 300 pieces are frequently not the same method. Almost every comparison table online gives you the first number and never mentions the second.
The setup charges are well established:
| Method | Setup / tooling | Runs on |
|---|---|---|
| Screen print | ~$20–50 per colour (one screen each) | Very low per-unit once running |
| Embroidery | $50–150 digitizing, one time | Per-unit is stitch-count driven, and does not go up with more colours |
| Heat transfer / DTF | Minimal — file prep only | Highest per-unit (film + ink + adhesive on every single bag) |
| Deboss / emboss | Die charge (ask your factory — it varies with size, detail and die metal) | Low per-unit |
| Metal plate | Mold charge, the largest of the group | Plate cost per unit, plus attachment labour |
Read the second column and the third column together and the whole thing reorganises itself:
- Four-colour logo? Screen printing needs four screens. That’s four setups before a single bag is printed, and screen printing can’t do a smooth gradient anyway (it fakes it with halftone dots, which break up on rough bag fabric). Stop asking for screen print quotes.
- Same four-colour logo, 300 pieces? Heat transfer wins on total cost, because it has almost no setup and 300 units isn’t enough to amortise anything.
- Same four-colour logo, 5,000 pieces? Now the per-unit cost of heat transfer is the dominant term and the setup is a rounding error. The ranking flips.
- Metal plate at 300 pieces? The mold can be a meaningful fraction of the entire branding budget. At 5,000, it’s noise — and the plate is a genuine asset you’ll reuse across every future style.
So the useful exercise is not “which is cheapest per bag.” It’s:
Total = (setup) + (per-unit × quantity)
Run it at your actual quantity, with your factory’s actual numbers, for your top two methods. It’s ten minutes of arithmetic and it regularly changes the answer.
One footnote with teeth: if you pay for a custom mold and hand it to the factory free of charge, US customs treats that mold as an assist, and its value has to be apportioned into the customs value of your bags. It’s a small, boring, five-year-look-back kind of problem. Ask the factory to invoice the tooling instead of “including” it. (More on why in our piece on choosing hardware that survives bulk.)
How will the logo actually fail?
Not by washing. Bags don’t get washed. Every durability claim you’ll read about logo methods — “holds up to 50 washes,” “won’t fade in the laundry” — was written for T-shirts and then copy-pasted onto bags by people who never thought about it. It is the wrong test entirely.
A bag fails in four ways, and each one attacks a different method:
| Failure mode | Where it happens | What it kills |
|---|---|---|
| Abrasion | Backpack base scraping walls, floors, overhead bins; tote corners | Screen print (scuffs), heat transfer (edges lift first) |
| Flex / creasing | Tote fold lines, soft-body bags stuffed and emptied daily | Screen print (cracks along the crease), heat transfer (peels at the fold) |
| Tension | Handle roots, strap anchors — anywhere load concentrates | Metal plates (rivets loosen), stitched patches (thread pulls) |
| Time / UV | The bag lives outdoors for two years | Deboss on soft PU (flattens back out), hot stamp foil (dulls), plating (tarnishes) |
Match the method to the failure mode of that placement, not to a general durability ranking:
- Logo on the base or lower panel of a backpack? That surface will be dragged across concrete. Screen print or a woven/embroidered patch. Not heat transfer.
- Logo on a folding tote? Anything ink-based will crack on the fold line eventually. Move it off the fold, or use a stitched patch.
- Logo on a handle root or strap? Don’t put a rigid metal plate where the load concentrates. It’s a stress riser and a rattle.
- Logo on a soft PU panel? Deboss will read beautifully in the sample and go quiet after eighteen months of the panel flexing. Firm leather or microfibre holds depth. Very soft PU does not.
None of that shows up in a sample photo. All of it shows up in your returns. If you want the actual test protocol — abrasion, rub fastness, the numbers a factory should be able to produce — that’s in bag durability testing.
Will your logo even be legible?
Embroidery needs roughly a half-inch minimum stroke width and one-inch text height to stitch cleanly. Screen printing goes down to about a fifth of an inch. More logos die on this constraint than on any other, and they die quietly, in the form of a sample that comes back looking blurry and nobody being quite able to say why.
Thread has a physical width. It cannot be thinner than it is. A logo with a hairline serif, or a tagline in 6pt, or a gradient, is not “hard to embroider” — it is not embroiderable, and the machine’s attempt to do it anyway is what produces that thick, mushy, slightly-wrong version of your wordmark that you can’t unsee.
Three rules that will save you a sampling round:
- Gradients and photographs: heat transfer or DTF. Not embroidery. Not screen print (halftone dots disintegrate on textured bag fabric).
- Fine text under about 1″ tall: print it, deboss it, or engrave it on a plate. Do not stitch it.
- Simple 1–3 colour mark at volume: screen print, and it will cost less than everything else by a distance.
Send the factory a vector file (AI, EPS, or PDF with outlines), with Pantone references, at the actual size it will appear on the bag. A JPEG pulled off your website is how you get a logo that’s technically correct and visibly wrong. The tech pack guide covers what a complete artwork package contains.
Where in the production line does each method happen?
Screen printing happens on flat cut panels, before the bag is assembled. Everything else can happen later. This sounds like a manufacturing detail. It is actually a deadline, and it’s the one that decides whether you can still change your mind.
- Screen print needs a flat surface and a squeegee stroke. Once the panel is sewn into a three-dimensional bag with seams, gussets and a zipper, you’re printing into a jig, only on the flattest available area, and a seam in the wrong place ruins the print. Which means screen print is a cutting-stage decision. Change it after cutting and you’re re-cutting.
- Heat transfer uses a small press plate that can be worked between seams. It survives being a late decision.
- Metal plates, patches, woven labels attach during assembly.
- Deboss and hot stamp are usually done on the panel or the patch before assembly, for the same flat-surface reason.
Practical consequence: if your artwork isn’t finalised, do not commit to a screen-printed logo on a cut-and-sew bag. You are locking the branding decision to the earliest possible point in the production schedule, and you will pay for that in re-cut fabric.
This is also why the logo belongs in the sampling conversation and not in an email three weeks later. Every method above has a strike-off — a test print, a stitch-out, a pressed sample on the actual substrate — and you want to see it on your real fabric, not on a swatch of something similar. On custom canvas totes this matters less, because screen print is the default and a strike-off costs almost nothing. On a structured nylon backpack with a coated face, it matters a great deal.
The decision, in order
Work down. Don’t skip. Each step legitimately removes options.
- What’s the face fabric? Cross out everything it fights. (Coated? Embroidery needs a laminate seal or it’s out.)
- How many units? Run
setup + (per-unit × qty)on your two finalists. Don’t trust the per-unit number alone. - Where on the bag, and what happens to that spot? Abrasion, flex, tension, or sun. Cross out what dies there.
- How many colours, how fine, any gradients? Cross out what can’t render it.
- Now pick the one you like best out of what’s left.
If step 5 leaves you with nothing, the problem is upstream — usually the artwork, occasionally the fabric. That’s a redesign, and it’s a much cheaper redesign now than in round four.
What to send us
- Vector artwork (AI / EPS / outlined PDF), at real size
- Pantone references
- The face fabric, or your best guess at it
- Placement, marked on a photo or sketch
- Quantity, honestly
- What the bag is for — a commuter pack and a beach tote fail in completely different ways
Send that and you’ll get back a method recommendation with the reasoning, not a menu. Tell us what you’re branding, or start with our printing and branding options if you want to see the range first. And if you’re still writing the brief, the RFQ checklist is the shortest path to a quote that survives bulk.
