The $5,240 Lesson: What a Bad Batch of AC Kits Taught Me About Buying on Unit Price

The 7:14 AM Call

It was January 2023. The 10th, if I'm remembering right—a Tuesday, snow on the ground, the kind of morning where you're still deciding whether coffee is worth leaving the car for. My phone rang at 7:14 AM. It was the procurement manager from our biggest automotive AC account, a regional service chain with 60 stores. She skipped the small talk.

"We need 200 air conditioning kits for cars. Delivered to our central warehouse by February 3rd. Full kits—compressor, fittings, strainer, the whole assembly."

I'd been handling HVAC and refrigeration parts procurement for four years by then. Knew the SKUs, knew the lead times, knew which suppliers picked up the phone on weekends. Or so I thought. What I didn't know was that this single order would cost my company $5,240 in avoidable damage—and permanently rewire how I evaluate quotes.

Three Quotes, One Bad Instinct

I sent the RFQ to three suppliers that morning. Two were established vendors we'd used for years. The third was a newer outfit—let's call them Vendor C—that had been aggressively courting us for a few quarters. They came back with a quote 22% below the others. On a 200-unit order, that was roughly $2,100 in apparent savings. Not pocket change.

The upside was obvious. The risk was less obvious—I kept asking myself: is $2,100 worth potentially missing the February 3rd delivery? We had buffer days. The other two quoted 18-20 day lead times; the deadline was 24 days out. Math said yes. Instinct said something else, but instinct doesn't show up on a savings report.

I approved the PO to Vendor C that afternoon. My notes at the time say: "Unit cost lowest by far, verifiable references, will monitor first shipment." I underlined "monitor." (Note to self: underlining the word is not the same as actually doing the thing.)

The order included copper press fittings, solder fittings as a backup option, drier strainers, and a mix of compressors—some dc compressor units for the hybrid retrofit kits and electric compressor assemblies for the newer EVs. Everything was specced in the PO. Diameters, thread types, voltage ratings, temperature ratings. All of it written down cleanly.

When the Pallets Arrived

Delivery landed on January 26th—one day early. That should have been my first warning sign. In my experience, the cheapest vendor is rarely the earliest.

We opened the pallets the same afternoon. Three problems surfaced in the first 40 minutes.

Problem one: the copper press fittings. We'd specified press fittings rated to 16 bar with a specific O-ring compound and dimensional tolerances per the manufacturer's spec sheet. What arrived looked right in the photos. In hand, the wall thickness was off. My QC guy put a caliper on ten random units—four were out of tolerance by more than 15%. For AC lines running R-1234yf, that's not a rounding error. That's a leak waiting for a hot July day.

Problem two: the drier strainers. Half the batch had a different mesh density than the spec. One of our senior techs caught it—the filter element was visibly finer in some units, which would restrict flow and could starve the compressor under load. Again, not catastrophic on day one. Very catastrophic on month six.

Problem three: the compressors. This one dropped my stomach. The order was split between dc compressor assemblies (12V and 24V variants) and electric compressor units. Vendor C had shipped the 12V dc units in the 24V boxes. Twenty-two units affected. If any of those had gone into a customer vehicle, the customer would have been looking at a burned-out compressor within weeks—and we'd have been looking at a warranty claim trail that led right back to our loading dock.

I only believed the advice everyone gives you—"verify before you ship"—after ignoring it once. $5,240 later, I'm a convert.

The Scramble

Here's where the story turns from bad to worse. We had nine days to the delivery deadline. I called Vendor C. They were apologetic, but apologetic doesn't restock a warehouse.

I called our two original vendors. One had partial stock on the copper press fittings and could get me the dc compressor units in five days—but not the electric compressor assemblies until the following week. The other could cover the drier strainers immediately but had no compressors on hand at all.

We split the order. Rushed the partials. Paid air freight on two shipments (which, honestly, felt excessive for parts that should have been correct the first time). Pulled two technicians off scheduled jobs for three days to inspect, repack, and relabel everything at our dock.

The order made it to the customer's warehouse on February 2nd—one day early, thanks to the scramble. They never knew anything had gone wrong. But I knew. And the invoice knew.

What This Actually Cost

Let me add up what that "22% cheaper" quote really cost us.

• Original PO to Vendor C: $9,410
• Replacement order from the two secondary vendors (partial air freight included): $11,880
• QC labor—3 days, 2 techs: $1,140
• Expedited repackaging and relabeling: $640
• Credit we gave the customer for a minor scheduling shuffle (goodwill): $1,200
• Vendor C refund (partial, after negotiation): -$3,200
• Disposal of unusable fittings and strainers: $170

Total cost: roughly $21,240. Original budget if I'd gone with either established vendor: roughly $16,000. Delta: about $5,240 in pure loss.

That's the number on the whiteboard in our team room now. Not because I want to relive it. Because I want every new procurement hire to see what "cheapest unit price" actually means in practice.

Total cost of ownership—not just the invoice, but the freight, the QC time, the risk exposure, the eventual rework, and the customer trust you spend to cover your own miss. It's not a theoretical framework. It's the difference between a $9,410 order and a $21,240 order on the same day for the same parts.

What I Do Differently Now

I built a pre-check list after this. It's boring. But since I started using it 18 months ago, we've caught 47 potential issues before they shipped. Four of those would have been six-figure problems on their own.

1. Sample first on new vendors, always. No exceptions, even when the deadline is tight. A $200 sample and a three-day wait beats a $5,000 rework every single time.
2. Spec sheets go on the PO in the vendor's own SKU language. Not ours. Theirs. If they can't map our requirement to their catalog reliably, they can't fulfill reliably either.
3. Budget for QC on every first order. If the quote doesn't include a line for inspection, add it yourself. It's cheaper than surprise freight.
4. Never split critical specs across multiple vendors without acknowledging the risk. Splitting worked for us this time—barely. It also added $1,780 in air freight and a week of lost sleep.
5. Ask for the last three orders with identical specs. If a vendor can't produce references for the same product family, they're guessing as much as you are.

There's a version of this story where I look smart and Vendor C is the villain. That's not the honest version. I had two reliable suppliers in hand. I overrode them for a discount I didn't need to chase. The failure was mine. The fix is also mine.

If you're sourcing copper press fittings, drier strainers, or compressor assemblies—especially splits between dc compressor and electric compressor variants—the unit price is the smallest number in the deal. The rest of the cost shows up later. Trust me. I have the invoice to prove it.