# How to know if your AI use case is actually worth building. The truth is, mos...
Canonical: https://social-archive.org/tgroenwals/ZHO2SiFUGw
Original URL: https://www.linkedin.com/posts/nadinesoyez_how-to-know-if-your-ai-use-case-is-actually-share-7467824510831001600-sian/
Author: Nadine Soyez
Platform: linkedin
## Content
How to know if your AI use case is actually worth building. The truth is, most AI projects are already headed for failure before anyone has even started. Because no one stopped to ask: should we even be doing this? Before I scope any AI project, I run four questions past the team: 1) Is there a real business problem here? Not "AI could probably help with this." A specific problem that costs you time or money, and keeps coming back. 2) Do you have the data to support it? AI needs clean, accessible, sufficient input. Most teams haven't checked this before committing to a project. 3) Can a human currently do this, just too slowly or expensively? If no human can do it reliably, AI won't either. That's the honest version of this question. 4) What does success look like in numbers? If you can't measure it, you can't justify it. And you won't be able to learn from it either. Teams nearly always skip straight to tool selection. These four questions come first. Run this before your next AI scoping conversation. It saves months of wasted effort. What question do you wish someone had asked before your last AI project started?