The "useful beats clever" test we run on every feature spec.
Four questions we ask before any new feature gets build time.
Clever features are tempting. They demo well and make the team feel good. But the features people use every day are usually plain and predictable. Before anything goes into a sprint, we run it past four questions.
1. Who uses it, and how often?
Name the person and the moment. "A parent, every evening, checking whether the kids finished their jobs" passes. "Users might like this" doesn't.
2. What do they do today instead?
If the answer is "nothing, and they're fine", the feature probably isn't needed. If it's "a spreadsheet and three phone calls", you're onto something.
3. What's the simplest version that solves it?
Strip it back until removing anything else would make it useless. Build that first.
4. How will we know it worked?
Pick one measure before you build: time saved, fewer support requests, more people completing a task. If you can't name one, you can't tell whether it was worth it.
What happens to features that fail
They go on a "not now" list with a note on why. Some come back later with better evidence. Many quietly disappear, and nobody misses them.
We use this test on our own products, like ChoreBuddies, and on client projects.