2026-11-11
Coatings Sourcing Problems Usually Start Before the Purchase Order
I’m Jane Smith, a senior content writer with over 15 years of experience in the packaging and printing industry. I specialize in writing about the latest trends, technologies, and best practices in packaging design, sustainability, and printing techniques. My goal is to help businesses understand complex printing processes and design solutions that enhance both product packaging and brand visibility.
The quote looked fine. The parts didn't.
I handle custom industrial coatings orders for a contract manufacturing group, and I've been doing it for nine years. In that time I've made six significant sourcing mistakes that I actually wrote down, and they cost us roughly $38,000 in wasted budget. I keep the checklist now mostly so I don't have to relearn any of it.
The one that still bothers me happened in September 2022. We placed an order for 1,800 powder-coated steel brackets. I collected three quotes, attached what I genuinely believed was a complete specification, and picked the middle option. The parts arrived on time and looked right. Six weeks after install, the coating started lifting along the punched edges. We stripped, re-coated, re-shipped, and ate the labor. That one came to about $6,400.
For almost two years I blamed the supplier. That was the wrong diagnosis, and it sent me down a path that wasted a lot more money than the original failure did.
What I thought the problem was
My first read was simple: we picked a bad coatings manufacturer. So the fix seemed obvious — build a wider supplier network, get more quotes, add a backup powder coatings supplier, negotiate harder on price and lead time. I did all of that. By April 2023 I had five qualified vendors instead of two.
The next failure came from a different vendor, in a different country, with a different process. Same symptom: adhesion loss at edges. That's when it stopped looking like a supplier problem and started looking like something we were doing.
What was actually going on
"Specification" isn't a shared language
Here's the line we were sending out, more or less verbatim: Powder coating, RAL 7016, 60–80 µm, matte. Internally, that felt complete. It has a color, a film thickness, and a finish. What else is there?
Quite a lot, it turns out. That line doesn't say how the substrate gets prepped. It doesn't say which gloss scale we're measuring on. It doesn't say who approves the color panel or under what light. It doesn't say what happens if the first article fails.
We weren't comparing three prices. We were comparing three different scopes that happened to share a color code. The cheapest quote was cheapest because it skipped a process step we never asked for — and never knew to ask for (note to self: this is the whole article in one sentence).
The line item nobody quotes
Pretreatment is where most coating failures live, and it's the thing buyers like me used to treat as invisible plumbing. ISO 12944 defines corrosivity categories from C1 through C5, and the prep standard that goes with each category is not the same. One of our suppliers was degreasing and phosphating. Another was blasting to Sa 2½ per ISO 8501-1. On paper, both were "powder coating." In a salt fog chamber, they were not remotely the same product.
I assumed "same specifications" meant identical outcomes across vendors. Didn't verify. Turned out every shop had its own default prep class, and every one of them considered that default obvious enough not to mention in the quote.
Test methods without pass/fail criteria
We used to write "500-hour salt spray" and consider ourselves rigorous. That phrase is basically meaningless on its own. 500 hours of what, tested per which method, on a scribed or unscribed panel, evaluated by whom, against what allowable creep from the scribe?
ASTM B117 gives you the apparatus and the conditions. It does not give you a pass/fail. That's your job as the buyer, and we were outsourcing it to whoever happened to write the report. One vendor's "pass" and our "pass" were about 3 mm apart at the scribe — which is the difference between a part that survives a season outdoors and one that doesn't.
The sourcing model confusion
This one took me longest to understand, and it's specific to how our side of the industry is structured. A coatings manufacturer owns the process. A coatings sourcing partner owns the relationship and the matching. Those are different jobs with different failure modes.
We asked a sourcing partner for a process-level guarantee they had no equipment to enforce. They said yes anyway (because who turns down an order), and we treated that yes as a manufacturing commitment. Honest mistake on both sides, but it was ours to prevent. When you buy through a sourcing layer, the guarantee you can actually get is about supplier selection and documentation — not about cure schedule on a line you've never seen.
What it cost
Adding it up honestly, because I think the number is the useful part:
- September 2022: $6,400 direct — strip, re-coat, freight, and about 40 hours of my team's time that nobody billed anywhere.
- April 2023: a $3,200 order with the wrong gloss level. We caught it at goods-in, so it was only $1,100 and three days. Lucky, not good.
- The language gap: I wrote "matte." They read "satin." When we finally measured, we were about 25 gloss units apart at 60°. Same word, two completely different surfaces. We'd been using it for years with other vendors who happened to share our definition.
- Q3 2024: a deadline with two hours left to decide. Normally I'd collect quotes and check references. There was no time. I went with the lowest bidder who said "about three weeks." We shipped nine days late, missed a customer's install slot, and paid $2,300 in expedite fees on our own line to claw back part of the schedule.
The direct costs are the small part. The expensive part is the second-guessing that follows — every subsequent order gets re-checked twice, timelines get padded, and the whole purchasing motion slows down for a year because one person got burned in front of everybody.
The part where I stopped optimizing for price
Here's the position I've landed on after six of these: when the deadline is fixed, an uncertain cheap option is the most expensive thing on the table.
In March 2024 we paid about $1,900 above the lowest quote to lock a written delivery commitment — with a stated penalty if it slipped. The alternative was a line stoppage on a $14,000 install. I ran the worst case twice before signing, because $1,900 felt like a lot for a piece of paper. It isn't. What you're buying is the difference between "probably on time" and "contractually on time," and after two rounds of probably, that difference turns out to be the entire product.
The same logic applies to coating quality, by the way. A guaranteed prep class and a named inspection method cost more per square meter than "we'll do it properly." Pay for the version with names in it.
What we do now
One page. Every RFQ goes out with it, and any quote that doesn't answer it line by line goes back:
The eight fields we refuse to send an RFQ without:
- Substrate and its condition (new steel, galvanized, previously coated, and if so, what with)
- Pretreatment class, referenced to a standard (ISO 12944 category, ISO 8501-1 grade)
- Coating system: product type, not just "powder" or "liquid"
- Film thickness in µm, with a stated measurement method
- Color and gloss, with the gloss number and the angle it's measured at
- Test requirements with pass/fail thresholds: adhesion per ASTM D3359, salt spray per ASTM B117, stated hours and allowed scribe creep
- First-article approval: physical panel, signed off by us, under a named light source
- Delivery commitment, in writing, with what happens if it slips
Four things made the difference, in order of how much they helped:
Send the same scope document to every supplier. Not a description — the same document, and require a line-by-line response. Half the "price difference" between quotes disappears the moment they're all quoting the same thing, and the other half is now explainable.
Approve a physical panel and keep it. Ours lives in a drawer with the approval signature, the date, and the light source we signed it under. Two color disputes have ended at that drawer without a phone call.
Write pass/fail into every test requirement. If a supplier can't tell you what failure looks like on their own report, they're not testing, they're documenting.
Ask what they can't do. The vendors we've kept are the ones who answer that question with specifics. The ones who say "no problem, we do everything" are the ones whose quotes have the most blank space in them.
None of this is exotic. It's a page, a drawer, and one uncomfortable conversation per order. The reason I wrote it down is that I'm the guy who needed the page and didn't have it, and I'd rather you skip that part.