I’m surprised by how many people think the value is in building the technology.
But building an AI project is like hiring someone for a role. The value doesn’t come from the hiring process. It comes from them doing the job, learning, improving, and delivering results over time.
Creating the AI tech is just the first step.
The real work is tuning it, so it drives better results. Making sure it supports the team the way they actually need support.
You also need a crystal-clear vision upfront. If this is a sales research agent, what’s it supposed to do: Save salespeople 15 minutes before every meeting?
Then the question becomes: is it actually happening? Are they still doing supplemental research, meaning we’re only saving 5 of those 15 minutes?
You have to talk to people.
Over-the-shoulder observation is also a great way to understand if you’re delivering what this was supposed to do.
Whatever the outcome - time savings, more iterations through a process, better decision quality - stay laser focused on it.
That clarity and continuous measurement is the 5% success zone.