How I Replaced a By-Hand Cataloging Job With a 30-Second Program

At the parts manufacturer where I work, someone had to tell the catalog which vehicles each product fits. By hand. One product at a time. There were more than 8,400 of them.

Power steering parts are fussy that way. A single kit might fit a few hundred different year, make, and model combinations. Get it wrong and a customer buys a part that doesn't fit their car, returns it, and trusts you a little less next time. So the matching matters, and for a long time the only way to do it was a person reading specs and typing.

When I looked at the actual process, the thing that stood out wasn't how hard it was. It was how consistent it was. The rules for deciding what fits what didn't really change from part to part. It was a job with a clear pattern, being done as if it were a typing job. Nobody had ever stopped to ask whether a person needed to be the one doing it.

So I built a program that does the matching itself. It takes the raw part information and the vehicle database, sorts each part into the right category, works out every vehicle it fits, groups all of that the way it needs to be grouped, and writes out the finished catalog in the exact format the industry runs on. The whole catalog, all 8,421 products, comes out the other side in about 30 seconds.

I want to be honest about the numbers, because it's tempting to inflate them. The old manual process was never timed, so I can't tell you I saved a specific number of hours. What I can tell you is that rebuilding the entire catalog used to be a project, and now it's a 30-second run that produces the same result every time. When something changes, I run it again instead of starting over.

Most of the worst manual jobs I find look like this one. Not complicated. Not interesting. Just a clear, repeatable task that a person got stuck doing because that's how it started, and nobody ever questioned it.