I'm a procurement manager at a 90-person building products distributor. I've owned our wood products budget — plywood, OSB, LVL, structural lumber, building panels — for seven years, negotiated with 8+ mills and distributors, and logged every order in our cost tracking system. That's somewhere north of $2M in cumulative spending on lumber and panels alone.
This isn't a brand brochure. These are the questions contractors, buying groups, and private-label buyers actually call me about — the ones I answer between meetings, with a spreadsheet open. Some answers are numbers. Some are judgment calls, and I'll tell you which is which.
Here's what's covered:
Four things, in this order.
Spec support. Can they issue a project-specific compliance letter — signed, referencing the actual standard — or do they just send you a catalog PDF and wish you luck?
Rejection behavior. When a load doesn't meet spec, do they take it back without a fight, or do they argue about whose inspector was right? This is the single best predictor of what the next 12 months will feel like.
Unit conversion. Can they requote your job on a per-thousand-board-foot basis using your actual cut list? Vendors who can't do this usually can't do a lot of other things either.
Delivery certainty. Last, but only because the first three filter out the noise. I don't have hard data on what percentage of mid-size suppliers blow delivery windows, but based on our own 7 years of order logs, my sense is it's higher than most people assume.
I reordered that list in 2022. Used to be price-first. Then a project got hung up on 7/16" vs. 15/32" OSB — a $300 material difference that turned into about $4,000 of labor rescheduling. Price moved to the bottom that week and it hasn't moved back.
You need four lines, and one question.
The four lines: thickness and tolerance, bond classification (Exposure 1 vs. Exterior), face and back grade (A/B, C/D, and so on), and span rating if it's structural. If it's a structural panel, confirm it's manufactured to an accepted performance standard — PS 1-09 for plywood, PS 2-18 for OSB. These are the standards your engineer is going to ask about, so get ahead of it.
The one question: "What thickness tolerance will be on the shipping documents?"
If the sales rep can't answer that, stop there. A tolerance number tells you whether the mill is holding the panel to spec or letting it drift to the outside of the range. And the difference between a tight and a loose tolerance shows up in your waste factor, not on the invoice.
One caveat: standards get revised. Verify the current edition before you cite it in a submittal — don't take a blog post's word for it, including mine.
More than a label. Less than most people think.
A real OEM or private-label program typically covers: custom lengths or end cuts, a proprietary appearance grade, your own wrapping and tagging, and often your own SKU structure. Sometimes it includes mill-direct packaging for a specific retail or dealer channel.
The mistake I made, in 2021: I assumed "private label" meant a spec change. It meant a label change. We got bags with our name on them and a length that was 2" off what our dealers wanted. Not the mill's fault — we never wrote it down.
What to put on the PO, explicitly: length, end treatment, grade, panel type, moisture content, labeling, and packaging. Every one of those. If it's not on the PO, it's not in the program.
Five years ago I wouldn't have looked. I do now.
Once you're evaluating a three-year supply agreement, company-level questions stop being abstract. Will there be a merger? Is a product line being exited? If a mill closes, where does the capacity go? Public governance disclosures — who's on the board, what they oversee, how the company talks about its segments — give you early signals. It's not a decision-maker on its own. It's one input.
Concretely: I keep a one-page supplier risk sheet. Ownership structure, segments, recent leadership changes, and board composition. Looking at Boise Cascade's board of directors took me about 15 minutes last quarter. That 15 minutes is cheaper than finding out mid-project that your EWP source is in the middle of a strategic review.
Treat it as context, not a verdict. Governance tells you how a company is likely to behave under pressure. It doesn't tell you whether their LVL will hold tolerance.
Naming collision. "Cascade" is a common word in Idaho business names, and "Boise" is a city — so search engines routinely blend company mentions together.
To be clear about what's what: Boise Cascade is a manufacturer and distributor in the engineered wood and building materials space. Cascade Auto Glass is a Boise, Idaho auto glass service business. They're separate companies in entirely different industries with no corporate relationship.
It matters for procurement because a diluted search result can send a new buyer down the wrong path — reading reviews for the wrong company, checking the wrong address, wondering why a windshield quote is attached to a lumber supplier. Quick check: look at the product catalog, not the company name. Catalogs don't lie.
Yes — on critical-path material. Full stop.
In March 2024, we paid about $600 extra for rush delivery on a late LVL beam package for a multifamily job. If I remember correctly, the alternative was holding an 8-person install crew for two days at roughly $2,400 per crew-day. The beams landed Thursday at noon instead of Monday.
What I bought wasn't speed. It was certainty. A vendor saying "probably Monday" is telling you they don't know. A vendor saying "guaranteed Friday 10 a.m., here's the confirmation number" is telling you they own it.
That said — the premium isn't always worth it. On non-critical material, take the slow lane and bank the difference. The question isn't "is faster better." It's "what does the delay cost me?" If you can't answer that in dollars within about five minutes, you probably don't need the rush.
Missed deadlines cost more than rush fees. That's been true every single time I've run the math.
Normalize everything to a per-used-unit cost, then add five rows.
I built our comparison sheet in 2020 after getting burned twice on quotes that looked clean and weren't. The rows: material, freight, spec-driven waste, rework or replacement, and downtime from late arrival.
Here's why it matters. We had two plywood suppliers about 22% apart on paper. Went back and forth for three weeks. On paper, the cheaper one made sense. But my gut said the cheaper panel was going to eat the difference in waste — so I added a column for thickness consistency and applied our actual cut plan.
That 22% gap compressed to about 4%. What I mean is: the real number was already close, and the cheaper quote carried more execution risk. We went with the higher-priced supplier because they'd put thickness tolerance in writing on the quote.
If a vendor won't commit tolerance to paper, model them at the worst end of the range. Most will negotiate. The ones who won't were never going to be a good fit anyway.
On-time percentage and defect rate, by supplier.
I didn't start logging those cleanly until 2023 — four years into managing a seven-figure annual wood budget. I don't have hard data on industry-wide defect rates, and I'd be skeptical of anyone who claims to. But based on our own records, my sense is that roughly one in ten first deliveries has at least one issue requiring a correction: thickness, grade, count, or labeling.
That number is only useful if you've been collecting it. If I'd started in 2019, I'd probably have renegotiated at least two contracts earlier and saved something in the range of $40,000.
If you're starting supplier comparisons today, one spreadsheet, six columns: date, supplier, line item, spec, receiving condition, resolution time. That's it. Two minutes per delivery.
One last thing, because it cost me. In 2021 I skipped written confirmation on a rush order — we'd used the same vendor for six years, so I thought, what are the odds? The odds caught up with me when the verbal agreement got lost in a shift change. Nobody's fault but mine. Now every order gets a confirmation email, even the ones I've placed 200 times. Especially those.