I've been handling custom packaging orders for B2B clients since 2018. In that time, I've made 11 mistakes big enough to document—wrong colors, wrong dimensions, wrong substrates—totaling roughly $8,400 in wasted budget. None of them were caught by "being more careful." They were caught by a checklist.
This is that checklist. Seven steps. Each one exists because I skipped it once and paid for it. If you order custom packaging 4-6 times a year and you're not a packaging engineer, this is for you. Running the whole thing takes about 40 minutes.
Step 1: Write the spec sheet before you talk to any vendor
In my first year (2017), I made the classic specification error: I told a vendor "standard box, gorilla boxer theme" and assumed we both knew what that meant. We didn't. I meant a sturdy shipping box with a gorilla-and-boxing-glove graphic. They printed a box featuring a literal gorilla wearing boxing shorts—the "gorilla boxer" design. The client had ordered it for a protein supplement launch. Nobody laughed.
That mistake cost $600 and taught me a rule: never describe a spec in words alone. Write it down, attach reference images, and make the vendor repeat it back to you before anything moves forward.
The spec sheet needs: dimensions, substrate, print method, color references (Pantone or CMYK), finish, quantity, and a physical sample request. If your vendor doesn't ask for this, they're guessing. Guessing is expensive.
Step 2: Separate the brand name from the product name
Here's a mistake that isn't mine, but I see it constantly in inbound requests: people search for "gorilla glue vape pen" when they want custom packaging for a vape product, and they end up on Gorilla Glue's website because the names overlap. Gorilla Glue makes adhesive. They don't make vape pens. But the search confusion is real, and it spills into packaging inquiries.
I've gotten three requests in the past year that started with "I saw gorilla glue vape pen packaging" and ended with us spending 20 minutes untangling what the client actually needed. The lesson: when you write your spec, strip out brand names that belong to someone else. Say "child-resistant vape cartridge box, matte finish, custom logo" instead of "like the gorilla glue one." Specificity saves time.
Checkpoint: if your request includes a brand name you don't own, rewrite it before you send it.
Step 3: Test your material claims against reality
We had a client ask for waterproof labels for outdoor equipment. I quoted a synthetic stock with a laminate, promised 18 months of weather resistance, and felt confident. Then the client said, "I want it to hold up like that gorilla waterproof patch and seal spray." That comparison stopped me cold. That product is a rubberized sealant—it bonds to a surface and stays put for years. A printed label is not a sealant.
I had to walk back my claim. The honest answer was: our label would survive rain and sun for 12-18 months, not "permanent." If I'd tested the material—literally stuck a sample on a piece of outdoor equipment and left it in the yard for two weeks—I would have known exactly what to promise.
Checkpoint: before you quote weather resistance, abrasion resistance, or adhesion strength, request a physical sample and test it under conditions that match the end use. Manufacturer spec sheets are a starting point, not a guarantee.
Step 4: Match the packaging to the archive life
Not every order needs to look good in a warehouse. Some need to survive 30 years in a climate-controlled room. I learned this from a library project—they needed archival boxes for a local history collection that was being digitized for the New Castle library catalog.
The librarian asked one question that changed my approach: "Will this box be opened 50 times over 20 years, or once and then stored?" The answer was "once, maybe twice." That ruled out our standard folding carton (too much wear at the hinges) and pointed toward a rigid chipboard option with a removable lid.
Most packaging buyers don't think about end-of-life durability. If your product is going to a library, an archive, a museum, or a collector, ask the question. It changes the substrate, the adhesive, and sometimes the print method.
Step 5: Verify dimensions with a physical object
I once quoted a custom box for a client who made replacement parts for American Flyer wagons—the vintage toy train cars. He gave me dimensions from a catalog: 8" x 3" x 2.5". I ordered the boxes. When they arrived, the parts didn't fit. The catalog dimensions were for the assembled wagon, not the individual parts, and nobody had caught it because I never asked him to send me an actual part.
That mistake cost $340 in discarded boxes and a two-week delay. Now, every order over $500 requires either a physical sample or a dimensioned drawing signed off by the client. Verbal or catalog dimensions don't count.
Step 6: Calculate total landed cost, not unit cost
Unit price is the number everyone compares. Total landed cost is the number that actually matters. I've watched buyers choose a vendor with a $0.12/unit advantage and then spend $400 on expedited freight because the production timeline didn't account for a holiday shutdown.
Here's a comparison that helps: people ask "how much does it cost to wrap a small car" and get quotes ranging from $2,500 to $6,000. The difference isn't just the vinyl—it's the prep work, the installer's labor hours, and whether the quote includes removal of old film. Packaging works the same way. The box is one line item. Tooling, setup, freight, and rush fees are the rest of the bill.
Checkpoint: build a simple total-cost table before you approve anything. Columns: unit price, setup/tooling, freight, rush surcharge, and estimated redo risk. The cheapest unit price rarely wins.
Step 7: Run a pre-press proof (and actually look at it)
I skipped this once in Q1 2022 and shipped 1,800 boxes with a typo in the phone number. That cost $1,900 in reprints and a client relationship I'm still repairing. The proof was sitting in my inbox for two days. I "approved" it from my phone without zooming in.
Now the rule is: one person checks the digital proof on a desktop monitor at 100% zoom, and a second person checks the physical proof under the lighting conditions the product will actually live in. Two sets of eyes, two formats, one sign-off document.
Common mistakes this checklist catches
- Specs in words only—no reference image, no sign-off.
- Brand-name borrowing—describing your product by referencing someone else's trademark.
- Unverified material claims—quoting durability without a physical test.
- Archive blindness—ignoring how long the package needs to last.
- Catalog dimensions—trusting a number without a physical object.
- Unit-cost tunnel vision—comparing box price while ignoring freight and redo risk.
- Phone-proofing—approving print files on a screen smaller than the final product.
A note on how this has changed: in 2018, I could get away with a phone call and a handshake on specs. In 2025, with more vendors, faster turnarounds, and shorter print runs, that approach gets you a reprint. The fundamentals—clear specs, physical checks, total cost—haven't changed. The tolerance for skipping them has.
Run all seven steps. It takes 40 minutes. The mistakes it prevents cost me $8,400 to learn.
