A Quality Inspector’s Guide: 6 Checks I Run Before Approving Any Order for My Job Shop

This Checklist Is for Any Shop That Orders Parts—Then Has to Use Them

I work in quality at a mid-size automotive parts supplier (think stamping, dies, the works). My job is to check what we order against what we need—before it hits the production line or a service bay.

This checklist came from years of doing that. It’s not a master list of every possible spec—it’s the six things I always check, in order, when a new batch of components lands on my bench. It’s saved me from more than one costly re-do. If you’re ordering parts for your own shop, or even just installing them at home, these same checks apply.

Step 1: Physical Fit Before Electrical Spec—Check the Connector Housing First

Everybody jumps straight to the electrical specs: voltage, resistance, dwell time. But if the connector housing doesn’t lock into the engine harness, none of that matters.

What I do: I take the part (say, a beru be6/1 spark plug connector 0300.111.00) and try to seat it onto the spark plug and the harness connector by hand. Not with tools. If I feel any binding or play, I flag it. That plastic latch is the most common point of failure in the field—I’ve seen it snap on an engine vibration after 200 miles.

Check point: The connector should click into place with a firm push. If it clicks twice (or not at all), that’s a problem.

Step 2: Surface Finish on the Ground Electrode—The ‘Fingernail’ Test

This one’s a bit old-school, but it works. For spark plugs—or really any precision ground component like the tip on a beru ignition coil—you check the surface finish with your fingernail.

What I do: Run your nail over the electrode. If you catch a ridge, it’s either a tool mark or a burr. That’ll cause premature wear and a weaker spark over time.

Why I care: A smooth finish means the part was ground with fresh tooling. A rough finish means the die was getting dull, and they pushed the batch through anyway. I’ve rejected entire orders over this. (Should mention: the tolerance here is usually Ra 0.8 µm or better—so the difference is almost invisible to the eye. You need the nail test.)

Step 3: Check the Muffler’s Internal Baffle Alignment—Not Just the Inlet/Outlet Diameter

Most people measure the inlet and outlet diameters of a muffler (makes sense—you need to know it’ll fit the exhaust pipe). But I always check internal baffle alignment. For a magnaflow performance muffler 13156, the internal straight-through design is what gives you the sound profile. If the baffles are offset by even a few degrees, the sound changes—and on some builds, that’s a fail.

What I do: Shine a light through the inlet and look dead-on at the outlet. You should see a clear, unobstructed path down the center. If the light doesn’t hit the center of the outlet, the internals are misaligned.

Real example: I caught a batch where the baffle was twisted 4 degrees (I only knew because I checked with a protractor—kind of a habit). The vendor said it was ‘within tolerance.’ We sent it back. Cost us a week, but saved a customer from a drone-y exhaust note they would have hated.

Step 4: Fuel Filter’s ‘Flow Direction’ Marking—Obvious, But Check the Angle

Every decent metal inline fuel filter has an arrow stamped on it. The arrow points from fuel tank to engine—that’s straightforward. What I look for is the angle of the arrow relative to the inlet fitting. If the arrow is stamped on a flat section offset from the fitting, a technician might mount it with the arrow straight up, but the line runs at an angle. That twists the filter housing and puts stress on the crimp seal. Over time (and we’re talking maybe 8,000 miles), that seal can leak.

What I do: I mount the filter on a mock-up bracket with the actual hose angles. If the housing twists >15 degrees under natural hose tension, I reject it. No one thinks about this until they smell fuel.

(I should add: a surprising number of filters don’t have a visible manufacturing date code either. That’s a separate issue—but if I can’t tell how old it is, I’m wary.)

Step 5: Serpentine Belt Routing—Verify Against the Actual Pulley Layout, Not the Diagram

I got a call from a service shop once: “How do you put a serpentine belt on a 2018 something-or-other? We have a diagram, but it doesn’t match.”

What I do: I always compare the belt routing diagram (printed on the fan shroud or an under-hood sticker) against the actual belt that came off. Not the manual. The actual physical belt. People take belts off and forget to note the routing—or they used a generic diagram from a database that’s close but not exact.

My check: I trace the belt path with a marker on a whiteboard before I remove the old one. If I’m approving a new belt, I make sure the width matches (obviously) and that the rib profile is correct. A mismatch by one rib (6-rib vs. 7-rib) can cause slip at idle—and that kills alternator charging.

Step 6: Last Step—Do a ‘Good’ vs. ‘Bad’ Side-by-Side

If I’m reviewing a new batch of parts (say, a beru ignition coil or a beru be6/1 spark plug connector), I pull one from the new batch and one from the last known-good batch. Put them side by side on my bench. I’m not looking at the specs—I’m looking for what changed.

Common differences:

• Parting line shift: If the mold or die is wearing out, the parting line will be slightly offset. You’ll see a thin ridge on one side.
• Surface finish change: A new batch might be shinier (different mold release) or duller (tool wear).
• Date code / logo clarity: If the mold number is harder to read this batch, it might be degrading. I’ve rejected parts for illegible marks before.

Common Mistakes I’ve Made (and Fixed)

I still kick myself for not doing Step 6 on a batch of fuel filters last year. The batch looked fine on paper. But side-by-side? The new batch’s housing crimp didn’t have the same knurled finish. That meant it was done on a different tool. That difference doesn’t always matter—except one time it did, and the seal leaked at 14 PSI. That cost us a $12,000 repair and a bruised customer relationship. Not a fun conversation.

So: that’s the list. Six checks, done in order. It worked for me, but my experience is mostly with stamped and formed parts for high-volume production. If you’re working with specialty one-offs or custom fabrication, the calculus is different. But if you’re ordering parts in any volume—or even just planning a single install—these steps are a solid start.