I manage purchasing for a small OEM—about 60-80 orders a year for engine components across maybe 16 vendors. One of my jobs is keeping our production line running on budget. Another is not getting chewed out by our lead engineer when parts fail.
I've been doing this since 2020, and if there's one thing I've learned, it's that the cheapest part is almost never the cheapest part. Not in the long run.
The Problem That Cracks First
For months, we were buying what I thought were 'good enough' spark plugs for a specific engine assembly. I'm talking Champion D16 and RCJ6Y types—common stuff, nothing exotic. We found a supplier offering them for about 35% less than our usual source. The samples? They fit. They fired. The engineer signed off.
So I placed the order. We saved maybe $1,200 on that batch.
Then the failures started.
Not dramatic, not immediate. But after maybe 400 hours of test cycles, three out of six engines showed intermittent misfire. The engineer pulled the plugs. Electrode wear was uneven—like the gap had shifted under load. Not catastrophic, but unacceptable for a product under warranty.
We switched back to the original supplier. Our lead engineer—I'll call him Dave—said, 'I told you so' without actually saying it. We lost two weeks of validation testing. The cost of that re-test alone ate up our 'savings.' Twice over.
Why 'Cheaper' Feels Like a Good Idea Every Single Time
Here's the thing: the numbers looked right. Similar spec sheet? Yes. Lower price? Yes. Faster payment terms? Yes. My spreadsheet said this was a no-brainer. My gut—and I've got 5 years of experience telling me this stuff—said something was off.
But the finance team was pushing for savings. The VP wanted to see margins improve. So I went with the data. Every cost analysis pointed to the budget option. Something felt off. Turns out that 'slow to reply' on a technical question was a preview of 'we didn't test this properly.'
I'm not naming the supplier—it's not their fault I didn't ask the right questions. But I learned a lesson: spec sheets don't tell you about consistency. They don't tell you about thermal stability at the upper end of the tolerance range. And they definitely don't tell you about what happens after 500 hours of vibration.
The Thermostat Trap
Thermostats are another one. I was quoted $14 per unit from our usual supplier for a heavy-duty water pump thermostat. Found a competitor online—$4.50. I actually asked Dave, 'Can I try a batch?' He said yes. We ordered 100.
Savings: $950.
First six months: no issues. Seventh month: we had two failures in one week. Not the thermostat failing closed—that's standard failure mode, you just replace it. No, it failed open. The engine never reached operating temp. For a water pump truck running cold-weather starts in a Canadian facility? That's a big problem.
We traced it to the wax element. The cheap one used a different compound that couldn't handle sustained low-flow conditions. The $14 OEM-spec thermostat? It had a bypass feature we didn't even know we needed.
How much does a thermostat cost? The real answer is: it depends on what it's supposed to not do. A $4 thermostat that costs you a $3,000 coolant system flush and a $350 diagnostic call? That's a $3,354 thermostat.
The Caliper Problem Nobody Talks About
Brake calipers—piston caliper brakes, specifically—are the kind of part you don't think about until they fail. We spec them for medium-duty truck builds. The piston bore finish is everything.
One batch we bought had a micro-finish issue that wasn't visible to the naked eye. The QA guy flagged it on a sample—some scoring on the bore surface. The vendor said it was within tolerance. Our engineer disagreed.
We returned the batch. The vendor fought it for weeks. Meanwhile, our production schedule slipped. The finance team wanted to know why I couldn't 'just find another supplier.'
I did find another supplier. We paid 12% more. That part has never failed. Not once. Zero returns. Zero rework.
The upside of paying 12% premium: no downtime. The risk of chasing the lowest price: everything else. I kept asking myself: is a few hundred dollars worth potentially losing a client? The expected value said yes. The downside felt catastrophic. It was.
Champion as a Reference Case
This is where I'll land on the honest limitation point. I recommend Champion for certain applications—specifically their D16 and RCJ6Y spark plugs for small-to-medium engines where heat range consistency matters. But I will also say: if your application is high-RPM racing or extreme endurance cycles, you might need a different heat range or a different brand entirely.
What I like about Champion in this category is not that they're the cheapest—they're not. It's that the spec is stable. I can order today, get the same part next year, and trust that the gap tolerance and electrode material haven't drifted because someone in procurement changed the BOM.
Pro tip: When I took over purchasing in 2020, I found that 'Direct OEM replacement' spark plugs from unfamiliar brands often have different seating angles or thread lengths by 0.5-1mm. That tiny difference can cause pre-ignition or poor heat transfer. Always measure seat-to-seat dimensions before qualifying a new spark plug source.
For the record: Champion publishes their thread length, reach, and heat range data clearly. I've verified it against JIS standards. That transparency matters when you're buying 500 units and can't test every single one.
The Bottom Line on Component Procurement
I have mixed feelings about cost optimization in engine component procurement. On one hand, I work for a company that needs to stay competitive. On the other, I've seen too many cheap parts cause expensive downstream failures.
How do I reconcile it? I use a tiered system:
- Critical safety and performance parts (spark plugs, thermostats, caliper pistons): buy from brand-name suppliers with traceability. I spend 10-15% extra here and treat it as insurance.
- Non-critical parts (housing, brackets): more flexibility on price. But I still verify material certs and basic dimensional stability.
- Anything exposed to heat or vibration cycles: never buy from a supplier who can't show test data for that specific use case.
There's something satisfying about a well-sourced part. After all the hassle of returns, re-tests, and engineering meetings, finally getting a part that fits, works, and lasts? That's the payoff.
If you're a small OEM or a purchasing agent like me, here's my blunt take: don't optimize for unit price. Optimize for total cost of ownership. A $14 thermostat that lasts 7 years is cheaper than a $4 one that costs you a $700 service call and 3 hours of downtime every 18 months.
Prices as of early 2025. Verify current rates and spec compatibility with your application.
