The wrong dishwasher part is a data problem, not a warehouse problem
My dishwasher died with a full day of kid-related dishes stacked in the sink, so I did the responsible-adult thing. I found the model number, matched the voltage and dimensions, confirmed the replacement pump was in stock, and paid twenty-five (!!!) extra dollars for overnight shipping so I could stop scrubbing forks by hand like a pioneer.
Three days later a box turned up with a pump that had three pins where my machine wanted four. It didn’t fit, and my dishwasher was still dead.
When the wrong thing lands on your doorstep, the instinct is to picture a warehouse: some poor soul grabbing the wrong box off the wrong shelf, a pick-and-pack robot having an off day. That’s almost never it. Distributors and manufacturers are astonishingly good at moving physical stuff. Where they fall down is the data about the stuff.
My part came off the right shelf. What was broken lived one layer up, on a product page that had stopped telling the truth months earlier and kept taking orders anyway.
Somewhere in my dishwasher’s model year, the manufacturer switched the pump from three pins to four and kept the exact same model number. The distributor pulled that supplier feed in once, published it, and moved on. Nothing in the system flagged the change, because catching it wasn’t anybody’s job until a customer (let’s be very clear: ME) paid to find the gap.
I emailed customer service, who caught my mild Canadian frustration first and decided (fairly) that this wasn’t their mess, so they handed it to the web team. The web team fixed that one SKU and called it done. The product team blames the supplier, the supplier blames the feed, and the same rotten data sits under a thousand other parts waiting its turn. Matt has a name for this: firefighting mode. You’re not maintaining a catalog, you’re running from one blaze to the next with a bucket.
The accountability actually sits higher up, with whoever owns commerce, because there was no gate. Nothing stood between a stale supplier feed and a live product that I could—and did—buy with my money. A little discipline catches the three-to-four-pin switch before I ever get to feel smug about my voltage research: a defined process, an automated check, a human spot-checking what the machine flags.
This isn’t a rounding error you can shrug off, either. Say it goes wrong half a percent of the time across hundreds of thousands of SKUs and hundreds of supplier feeds. That’s a steady drip of people getting the wrong part, and businesses getting a whole pallet of the wrong part.
It’s also why, in a lot of industrial categories, only ten to fifteen percent of revenue runs through ecommerce. Buyers aren’t allergic to self-serve. They tried it, got burned, and now the contractor who ordered the wrong part once will phone a rep and text a photo before he trusts the website again, which turns your expensive salesperson into a very slow search bar.
The fix is unglamorous and it’s a habit rather than a heroic cleanup. Matt compares it to fitness, which frankly feels like a personal attack, but I digress. You don’t go to the gym once and coast on it for life. You define the process, automate the parts that are chronically understaffed, keep a person in the loop to check the work, and you do it again next week, until keeping product data current is just how the place runs.
My dishwasher stayed dead for a few more days. The lesson is still very much alive and it scales to every catalog on the planet: the wrong part on the doorstep is the last symptom of a problem that started way back upstream, the day someone loaded a feed and assumed the job was finished.
Listen to Episode 1 of Data vs. Commerce wherever you get your podcasts.
