The boring feature always wins

There's a business I work with — a grocery wholesaler in Nyeri. Two shops, a stack of corporate clients, hotels and hospitals buying in bulk on credit.

When I sat down with the owner, the exciting problem jumped out right away. She runs everything on WhatsApp. Orders come in on WhatsApp. Invoices go out on WhatsApp. She chases payments on WhatsApp.

So obviously — obviously — the move is to automate WhatsApp. Hook up the API. Let the system send the reminders, take the orders, fire off the invoices. Magic. I got excited. I started scoping it.

And then I almost made the classic mistake: I almost built the impressive thing instead of the useful thing.

Because here's what was actually going on. This woman was running a real, growing business out of her own head. Each client's prices? In her head. Who owed what? Her head. What was in stock at each shop? Her head. Whether the business was even making money? She honestly couldn't tell you.

She wasn't missing automation. She was missing a backup of her own brain.

The WhatsApp robot was the shiny object. The gold was the most boring software you can picture: a list of clients. A number next to each one for what they owe. A button that makes an invoice. A screen that shows what's in stock. A ledger. I was about to skip the ledger to go build a robot.

So I stopped and built the boring stuff first. The WhatsApp “automation”? In version one it's a button that writes the message so she can paste it in herself. Took an afternoon. No API, no three months. And it does about ninety percent of what the fancy version would've — because the hard part was never sending the message. It was knowing who to send it to and what it should say.

The exciting feature and the valuable feature are almost never the same feature.

The exciting one demos well. The valuable one is usually a list, a number, and a button — something so unglamorous you're a little embarrassed to call it the centerpiece.

Build the embarrassing version first. Ship the ledger. You can always add the robot later — and half the time, once people have the ledger, they stop asking for the robot.

The boring feature wins. It just never gets to give the cool demo.