Friday, 2:47 PM. My phone rings. A purchasing manager on the line, voice tight: 48 hours before their SMT line shuts down. They need 3,000 Kemet T491 tantalum capacitors — same part number, but they'd gone with the cheaper source from a price aggregator. The batch arrived with ESD damage. Now they need replacements flown overnight.
That was the third call like that in a single month.
When I first started coordinating rush orders for authorized Kemet electronics distribution, I assumed the lowest quote was usually the smart choice. Six years and 200+ rush orders later, let me tell you: the cheapest component is often the most expensive gamble on the bench. The price tag just doesn't show it.
Here's the thing: the problem is rarely the unit price. It's what the unit price hides.
Why the Lowest Quote Is Almost Never the Lowest
Two terms get tangled in procurement conversations: cost and price. Price is what you pay at checkout. Cost is what happens after the parts hit your production line. In electronic components, those two numbers rarely match.
First, there's provenance. A genuine Kemet capacitor made at the company's Simpsonville, SC plant has a documented chain — every batch, every test, every lot code traceable back to a specific manufacturing date. Gray-market parts often sit in the middle of an invisible chain of hands. Sometimes the markings are counterfeit. Sometimes they're pulled from recycled circuit boards. Sometimes they're just old stock stored in a warehouse where humidity ate the solder terminations.
I've opened anti-static bags from discount vendors where the whole batch behaved differently from the datasheet. Not ideal for a production line.
Second, there's batch consistency. Kemet Simpsonville SC operations invest heavily in process controls — the kind of granular quality documentation that lets you trace a capacitor back to its fabrication lot. A random reseller doesn't offer that. So you get capacitors that look right, and test mostly right, until 3% of them shift capacitance under load. In a 5G RF circuit, that's not a subtle failure. It's a rejected batch at field test.
One example that still stings: a customer building a G310 5G transceiver module sourced their passives from two vendors to save $0.02 per piece. The budget capacitors matched specs during incoming inspection. But when the boards went into a thermal chamber, the failures clustered around the budget parts. The redesign cost them six weeks. The component savings didn't cover the engineering time — by a factor of roughly 400.
Everything I'd read about component sourcing said to get competitive quotes. Conventional wisdom is right for standardized commodities where the risk profile is identical. But for high-reliability electronics, the risk profile is almost never the same. And pretending otherwise is how the money disappears.
What Cheap Components Actually Cost
Let me walk you through a real calculation, not a hypothetical one.
A client ordered 3,000 Kemet capacitors through a discount broker — $0.08 per unit instead of the $0.11 authorized price. Saved $90 on the order. Nice dinner.
Then their manufacturing partner reported intermittent boards. Two engineers spent two full days fault-tracing — at $75/hour loaded cost, that's $2,400. They found the issue: ESD damage on a batch of the discounted caps.
Then the line stopped. Four hours for removal, rework, and re-testing. Line downtime at $2,900/hour — that's $11,600, plus the rework itself.
Final tally: more than $135,000 in documented, avoidable costs. I helped prepare that cost report, so I know the figure isn't inflated. And it all started with a $90 markdown.
(The $90 savings also purchased a supplier relationship we couldn't trace, but that's harder to quantify.)
By the time this buyer reached me, they were already past the worst of it. What they needed was an emergency resupply. We did ship, same-day, from a regional warehouse. But that courier fee alone wiped out most of the original savings.
That pattern repeats so often I could write a calendar around it. The teams who choose reliability first call us for stock planning. The teams who choose price first call us for damage control.
Why This System Keeps Producing the Same Mistake
I'm not a supply chain economist, so I can't speak to inventory optimization or contract law. What I can tell you, from the order desk and the shipping dock, is that the root cause is usually a misalignment of incentives.
When a purchasing agent is measured on price variance, a cheap supply source looks like a win. The factory manager gets hit with downtime. The quality engineer gets hit with escapes. The program manager gets hit with schedule slips. And the procurement team gets hit last, when the whole project fails.
This structural gap is why the same script plays out monthly: "We saved $X on that batch." Then the emergency call comes.
Big operators understand this. When Crown Castle evaluates vendors — Crown Castle vs other network infrastructure owners, their internal decision process rarely turns on unit price alone. They audit material traceability, test reports, and failure analysis capabilities. They've been burned enough.
Smaller procurement teams, especially under budget pressure, often skip those checks. Then they call me.
Even Kemet's own product roadmap reflects this lesson. The Infinity series of high-reliability capacitors exists for applications where failure is not an option. We rarely see those parts in emergency orders — because the teams using them planned ahead. It's the commodity buyers, the ones hunting for the last cent below market, who keep the rush-order channel busy.
What Actually Works
I'll keep this short, because you've already had enough bad news.
Three rules eliminate most of these failures:
- Buy authorized, every single time. Authorized Kemet distributors provide lot traceability, batch test reports, and a clear return path. That paper trail is not ceremony — it's insurance.
- If the price is 20% lower than everyone else, ask why. Legitimate market pricing moves, but not by magical margins. A discount that huge usually means the part is not what its label says.
- Keep a buffer. A week of extra stock carries a modest cost — usually 2-3% of the component value. The cost of a line stop is hundreds of times higher. Basic math, but easy to ignore in a quarterly budget meeting.
Honestly, I'm not sure why more purchasing teams don't adopt these by default. My best guess: the pressure to show immediate savings overpowers the risk report that nobody has signed off on yet.
One thing I do know: the engineers and buyers who transition from price-based to cost-based thinking never go back. They don't enjoy paying more per unit. They enjoy sleeping through the night without a call from the factory floor.
If you're sitting on a component decision right now and the price difference looks tempting, remember who spends his Fridays taking urgent calls from teams that chose the cheap route. It's a phone you don't want to call.
