Why We Picked the 'Cheaper' Material — And Why It Nearly Cost Us a Client
In 2022, our quality team pulled me into an emergency meeting about elastomer seals we'd been running for an automotive HVAC component. The seals were six months old — silicone ones had never failed early. Now, roughly 12% of eight production batches showed cracking.
The customer held back about $42,000 in payment. One order.
I'm not going to pretend this was a supplier problem. I made the call. After a cost review in 2021, I switched our supplier from silicone to nitrile rubber because nitrile ran about 38% cheaper per kilogram. On paper, we'd save $24,000 a year.
I thought I'd negotiated a win. But I'd been managing this category for nearly seven years — roughly $1.8M in annual elastomer spend — and I should've known better. Material substitution isn't like switching a box supplier.
The Real Problem: We Were Comparing the Wrong Numbers
Most procurement teams compare nitrile vs silicone the way they'd compare commodities. Price per kilogram, lead time, minimum order. Plug it into a spreadsheet, pick the lower number.
Elastomers aren't commodities. Their specs are performance envelopes, and each material behaves completely differently under stress. Silicone handles roughly −60°C to 200°C+. Nitrile sits closer to −40°C to 120°C — and near the top of that range, its aging curve steepens fast.
Our automotive customer's HVAC system peaks above 110°C. The nitrile held up near that edge. Silicone held up in the middle of its comfort zone. We looked at a datasheet that said both were 'suitable.' What I didn't understand then: supplier 'suitable' and production-line 'suitable' are different things.
And here's the second part, which still frustrates me — nobody flagged that our silicone molds had been designed for silicone shrinkage rates. Nitrile shrinks differently. Running nitrile through our existing tooling pushed dimensional tolerances right to the edge, and about one in eight batches showed weak sealing at the lip.
Same story with silicone overmolding. Some of our parts bond rubber to metal inserts, and nitrile and silicone behave completely differently there. Silicone bonds through a primer chemistry we'd already qualified. Nitrile needs a different primer system, one we didn't have. Different cure profiles, different adhesion failure modes.
If I'd priced all of that in upfront, the 38% per-kilo savings evaporates. The per-part cost should include tooling adjustment, primer processing for overmolding, and scrap rates — not just price-per-kilo × volume.
The Costs Nobody Flagged Before I Signed
Let me run the actual math from our 2021–2022 audit. These are audited numbers, not estimates.
What we thought we'd save: $24,000 (based on the 38% per-kilo gap).
What actually happened:
- Overmolding primer retrofit: $6,800 — new primer line and process qualification.
- Tooling adjustment: $4,200 — shrinkage mismatch required mold corrections.
- Scrap and rework: $11,300 — that alone ate up half our projected savings.
- Customer chargeback: $42,000 — preventable.
- Expedited freight: $3,900 — air shipments to cover replacement lots.
Total: $68,200. The 'savings' actually cost us an extra $44,200 — about 1.8× our expected savings.
There's another cost I took longer to account for: time. Two engineers spent six weeks on root-cause analysis. I lost three weeks on customer quality calls. Our procurement team spent five weeks re-sourcing. I won't pretend to price that precisely, but it's not nothing — maybe $20,000 in loaded labor if I'm being honest. That 'savings' never had a chance.
I should mention one thing: the overmolding primer retrofit cost was specific to our parts because they bond to metal inserts. If your parts don't have inserts, that line item doesn't exist.
What We Changed Afterward
Late 2022, we rebuilt our elastomer supplier evaluation. It took a few days. Nothing exotic.
First, we removed 'price per kilogram' from the decision criteria entirely and replaced it with cost-per-part-per-condition. Instead of asking 'what does this material cost per kilo,' we now ask 'what's the full lifecycle cost per part at continuous 110°C service.'
Second, we added a process-validation gate to every new supplier. Not a datasheet review — a small-batch qualification run where we measure scrap rate, dimensional stability, and overmolding yield before we even talk price.
If you're running silicone molds for overmolding, I'd suggest writing primer compatibility into the RFQ. We didn't, so our supplier quoted us a lower number that didn't account for an extra process step. That wasn't their fault.
One small thing that saved us later: when you're buying silicone through distribution, matching the shin etsu logo and batch number against the datasheet takes about thirty seconds. We caught one counterfeit lot in 2023 that way — mechanical properties were way off, and the price had looked 'too good.'
Not Every Application Needs Silicone — And That's Fine
I want to be clear: this isn't an argument for silicone everywhere. If your service temp stays under 80°C, your media is ordinary mineral oil, and you're not chasing tight tolerances, nitrile works fine. It genuinely saves money.
Some of our low-pressure fuel-line seals still run nitrile today. Low temp, steady pressure, five years of service, no issues. That 38% per-kilo savings is real in that context.
But if you're dealing with high heat, thermal cycling, UV exposure, or medical/food-contact requirements, nitrile isn't in the conversation. Not a pricing issue — a physics issue.
We moved back to silicone for our high-temp HVAC and electronic overmolding parts. We evaluated several suppliers and ended up qualifying Shin-Etsu Silicones of America's product line. Not because it was cheapest — it isn't — but because batch-to-batch consistency showed up clearly in our overmolding yield. Same tooling, same parameters, and scrap dropped from around 6% to under 1.5%.
That said — that result came from our specific overmolding process and tooling. If you're running straight compression-molded parts, or your tolerances aren't that tight, other suppliers' materials might fit better. My experience is based on roughly 200 elastomer orders, almost all in automotive and industrial electronics. If you're in medical or consumer segments, your calculus is different.
Looking back, my mistake was treating per-kilo price as the only variable. The real driver of total cost is whether a material runs stably through your entire process chain. Everything else is noise.
"The cheapest material isn't the one with the lowest quote. It's the one that doesn't force you to re-engineer your process to make it work."
If you're in the middle of a nitrile vs silicone decision right now, run one small qualification batch before you commit. It takes two weeks. It's cheaper than a chargeback.