🌮 Staring at a Brick Wall

Back in 2000-something working at the DMV, I was responsible for deploying the quarterly software updates to the 70 or so field offices in South Carolina. This software was responsible for everything having to do with obtaining a license. Literally everything, from taking the license photo to entering traffic tickets and looking up your license status.

So things were hectic as always, and we had to push these updates after hours because it was the DMV and you couldn’t shut it down. It’s 5:30pm and the last office is almost closed, so we start pushing the update. Everything looks fine, until the testers start testing the now-complete update at one of our field offices.

When they went to take a picture, the whole thing crashed.

If you’re familiar with the DMV anywhere in the US (or world most likely), you know they take a LOT of photos. If we can’t take photos, we’re cooked.

We re-create the problem at HQ - no problem there. Photos process fine. We re-push the update to the field office. Crash. Change some code. Crash. Replace the PC at the field office. Crash. Replace the camera. Crash. Everything was coming up crash.

Here’s some trivia for you - at every South Carolina DMV field office, there’s a wall with a little blue fabric backdrop attached to it. That’s where people stand to get their photo taken. In this specific field office, the blue cloth had fallen off at sometime during the day. So the camera was dutifully taking test photos of a brick wall when it crashed.

I remember one of our COBOL guys finally figured it out around 1am with this gem of a quote:

“Why the f*** are you taking a picture of a brick wall? Put a effing person in front of the camera!” We were all very tired.

Carl was right. Why would we ever take a picture of a brick wall? There’s no reason to test for this, there’s no reason any DMV would photo the brick wall. Very obvious in hindsight. I put it into our testing guidelines shortly after, and we stuck in some code to deal with larger image sizes, just in case.

Anyway, we all learned about edge cases that night, and the most important thing we did was write it down and learn from it for future software updates.

Here’s the seamless transition.

I talked in an earlier post about paying attention to customers’ experiences with your brand and how to capitalize on those. In that same vein, you should also be tracking metrics of content you push out for your prospects.

Focus on the content that gets replies, likes, shares, and more traffic. Early on you won’t have much to go on, but as you start to hone your craft and get more out there, you’ll have more data to take action with.

Just don’t leave your fans staring at a brick wall, lol.

Get outta here, go make something.