The Problem You Think You Need to Solve May Not Be the Right One
The best AI opportunities don't always start where you expect them to.
When organizations begin exploring AI, they often come to the table with an idea already in mind.
Could we automate this? Could AI handle that? Could we build something that makes this process faster?
Those questions are valuable. They get people thinking differently about how work gets done and what's now possible.
But they can also lead organizations to jump from problem to solution too quickly. The first problem you identify isn't always the problem worth solving.
Start With the Work, Not the Solution
It's easy to look at a frustrating process and immediately start imagining an AI solution for it.
A better place to start is understanding the work itself.
How does the process actually happen today? Who touches it? Where does information come from? Where does it go next? Which steps require judgment? Where are people creating workarounds just to keep things moving?
And, most importantly: Where is the real friction?
The answers aren't always obvious from a process map or leadership conversation. Often, the people doing the work every day have an entirely different view of the problem.
That's why getting the right people in the room matters.
Get Closer to the People Doing the Work
Your people know where the friction lives.
They know which task takes 45 minutes even though it seems like it should take five. They know which information has to be entered twice. They know which spreadsheet everyone quietly depends on, where communication breaks down, and which process only works because someone has figured out a workaround.
Those details matter when evaluating AI opportunities.
Instead of asking only: “Where could we use AI?”
Ask: “What's getting in the way of better work?”
That shift can lead to a completely different opportunity.
The First Idea Can Lead You to a Better One
We've seen this happen in our own work.
A team came to us with a specific idea for how AI could help solve a problem they were experiencing.
It was a reasonable idea.
But rather than immediately building around it, we got into the actual workflow with the people doing the work. As we learned more about the process, we uncovered another opportunity—one with greater potential to improve the way the work was getting done.
The original idea wasn't necessarily wrong.
It simply wasn't the best place to start.
That distinction matters.
When organizations rush to implement the first AI opportunity they see, they may spend time and money optimizing something that doesn't create much meaningful value.
Sometimes the better investment is a few steps away.
Evaluate the Opportunity, Not Just the Technology
AI can do a lot.
That doesn't mean everything it can do is worth doing.
A promising opportunity should make sense beyond the technology itself.
Does it address a meaningful business challenge?
Will it create capacity?
Will it improve how work gets done?
Does it fit within the organization's existing processes and technology?
Will the people who need to use it actually use it?
Can it be implemented responsibly and securely?
And ultimately: Is the impact worth the investment?
Those questions help separate interesting AI ideas from meaningful business opportunities.
Knowing What Not to Build Is Part of Good AI Strategy
There's a lot of pressure right now to do something with AI. That can make progress feel like the number of tools adopted, experiments launched, or solutions built.
More AI doesn't automatically mean better business.
Sometimes the smartest recommendation is to build.
Sometimes it's to teach your people how to use something that already exists.
Sometimes it's to improve the process before adding technology.
And sometimes it's to leave an idea alone and invest somewhere else.
Good AI strategy creates the confidence to make those distinctions.
Find What's Worth Pursuing
You don't need to come to the table knowing exactly what you should build.
Bring the business challenge. Bring the process that's frustrating your team. Bring the AI idea everyone keeps talking about.
Because before asking: “What can we build with AI?”
It may be more valuable to ask: “What's actually worth changing?”
