I stopped buying the "best" O-ring, air hose, and silicone sealant three years ago. Here's what I buy instead.

I'm an office administrator for a roughly 200-person company across three locations. I manage the MRO ordering for our maintenance side — about $60,000 a year across eight vendors, give or take. It's not glamorous work. But it taught me something that a lot of procurement advice gets backwards.

There is no "best" O-ring, air hose, or silicone sealant. There is only the product that matches your specific failure mode. If you're buying "the best" without knowing what's going to break first — you're buying insurance against a risk you haven't identified.

I learned this the expensive way. In 2022, when I took over the MRO spend, I tried to standardize everything on "premium" products. Fewer SKUs, higher quality, right? By the end of that year, I'd blown about $4,300 on over-spec'd parts that solved problems we didn't have. The maintenance manager was polite about it. His face was not.

O-rings: Buna-N is fantastic until it isn't

So — what is an O-ring used for, anyway? Simple answer: sealing a fluid or gas where two surfaces meet. That's it. But that simple job is exactly why buyers get it wrong, because the failure mode depends entirely on what's pressing on it.

Take oil-resistant Buna-N O-rings. Great product. Excellent for petroleum-based oils and greases, decent temperature range (roughly -40°F to 250°F), good mechanical strength, and cheap enough that you can stock a full AS568 size kit without blowing your budget. For hydraulic and pneumatic equipment in our shop? That's most of what we buy.

But here's the thing I walked into. People think oil resistance is the universal answer. Actually, Buna-N fails fast when the real threat isn't oil — it's ozone, UV, and weathering. We had a compressed air line running along an exterior wall. Buna-N O-rings in the exposed fittings started cracking within about six months. Not because they were "bad." Because I bought for oil resistance when the actual failure mode was sun and ozone. That's what EPDM is for.

I assumed "oil-resistant" meant "durable." Didn't verify. Turned out it meant "durable in one specific direction." Learned that lesson about $1,800 in replacement parts and one very annoyed maintenance tech.

So here's my actual recommendation, and I'll say the quiet part out loud: if you're buying oil-resistant Buna-N O-rings for an outdoor application, you're buying the wrong O-ring. Not because it's bad — because it's solving for the wrong failure mode. If your application is indoor, oil-facing, and moderate temperature, it's probably the best value per seal you can buy. If you're not sure, spend ten minutes identifying what's most likely to kill the seal first. That ten minutes beats a $900 emergency order every time.

Silicone sealants: 3145 is great — but 888 is not a "better" version

Same problem, different aisle. We stock Dow Corning 3145 RTV silicone adhesive for general sealing and bonding on equipment that sees moderate flex and temperature swings. It's a solid, forgiving one-part RTV. The data sheet gives it around -60°F to 400°F service range, though I generally don't push it past 350°F in anything we own. I'm not 100% sure of the exact number on the current TDS, so check it before you rely on it.

Now — here's where people go sideways. They hear "Dow Corning 888 silicone joint sealant" and assume it's the upgraded version of 3145. It is not. It's a different product for a different job. 888 is a construction-grade joint sealant — for expansion joints, curtain wall, weather sealing on buildings. It's designed for high-movement building joints, not for sealing a pump housing or a sheet-metal seam.

The assumption is that higher price means better performance. The reality is that price tracks the application the product was engineered for, not a universal quality ladder. If I use 888 on a $40 fitting, I'm not buying more durability — I'm buying a joint sealant where a general-purpose RTV would have been fine, and I'm setting myself up for a sealant that cures with less flexibility than the failure mode requires. Wrong tool.

My honest rule: if the joint is indoors, low-movement, and doesn't need to survive weather, 3145 is closer to what you want. If it's an actual building joint that moves with temperature, 888 is the right call. If you're not sure which category you're in, you don't have enough information to place the order yet.

Air hose supplier: bigger is not better — closer is better

Different aisle, same lesson. When I consolidated vendors in 2024, I moved our air hose supplier from a national industrial distributor to a regional one about 40 minutes away. I assumed the national distributor would win on price. I was wrong.

It wasn't a price issue at all. It was a stock and lead-time issue. The national supplier could get me most hose sizes in 5–7 business days. The regional supplier stocks the 3/8" and 1/2" IDs we burn through most, and gets anything else within 48 hours. When a compressor line goes and the shop is dead, 48 hours is the difference between an inconvenience and a lost production week.

Don't get me wrong — the national distributor was fine. To be fair, their catalog was deeper and their pricing was competitive for what we order. But "deeper catalog" only helps if I'm ordering something unusual. Ninety percent of our air hose orders are boring, predictable, and urgent. I need the boring, predictable, urgent items on a shelf within driving distance. That's it.

What I'd tell another admin: pick an air hose supplier based on how fast they can get you the product you buy most often, not the product you buy once a year. The once-a-year stuff can wait. The twice-a-month stuff cannot. If a supplier's strongest selling point is "we have everything," ask them what they stock locally. The answer usually tells you more than the catalog email.

"So you're saying buy whatever's cheapest?" No — that's not it.

I get why this reads that way. If I'm not buying "the best," and I'm not buying the cheapest, then what? The answer is: buy for the failure mode, and pay for over-spec only when you can name the specific failure you're insuring against.

For safety-critical equipment, or for anything where a failure takes the shop offline, buy the better-spec'd part. Paying 40% more for a seal that won't fail at 2 a.m. is the cheapest thing you'll do all quarter. That's not a contradiction — that's matching cost to consequence.

For the rest of the shelf — the fittings, the general-purpose RTV, the standard hose sizes — the "best" is usually the one that fits, arrives on time, and doesn't require a phone call with a customer service rep who has to escalate three times.

It took me about three years and probably 150 orders to figure this out, and I still get it wrong sometimes. But the question I ask before any MRO order now is simple: what's going to break this, and does this part solve for that break? If I can't answer the first half, I don't buy. That one habit has saved my department more money than any negotiation I've ever done — and it's the closest thing to a "best product" strategy I've ever found.