Before Your Next AI Idea, Ask This One Question
At JB Sales, we've caught the AI bug. The ideas don't stop. That's not a complaint -- most of them are actually good. But somewhere in the last few weeks, I started noticing a pattern that was slowing us down more than speeding us up.
The question we kept asking was "what can we do with AI?" Easy question. The answer is almost always: a lot. The harder question, the one we hadn't been asking consistently, is "what should we do with AI?" Those two conversations look similar on the surface. They're not.
The filter is outcomes
When an AI idea comes in, the instinct is to evaluate it on its own merits. Is it clever? Does it solve something annoying? Could it save time? Those aren't bad questions. But they're the wrong first questions.
The right first question is: what business outcome does this support?
At JB Sales, we have real goals. Things we're actually trying to move. When I started running every AI idea through that filter, a lot of them didn't make it. And that used to feel like saying no to progress. It doesn't anymore.
Cool isn't a business outcome
This is the thing I had to say out loud before it actually changed anything.
Cool AI work is genuinely satisfying. Building an elegant workflow, automating something tedious, finding a tool that does something you didn't think was possible -- there's real energy in that. I understand why teams chase it. But if the work doesn't connect to something the business is trying to accomplish, it's a detour. A well-intentioned one, maybe. Still a detour.
The shift is treating outcomes as the first question, not a box you check at the end to justify the work you already did. Before we get excited about what AI could do, we ask what we're trying to achieve. Then we look at whether the idea actually moves us toward that.
Why this is harder than it sounds
The ideas come fast. They're persuasive. Some of them come from smart people who are genuinely trying to help. Saying "that's interesting, but it doesn't connect to where we're going right now" takes more discipline than it sounds like in theory.
There's also a subtler problem: if you don't have a clear outcome to filter against, the question doesn't help you. "Grow the business" isn't specific enough. You need to know what you're actually trying to move -- specifically enough that you could evaluate an AI initiative against it and get a real answer.
What happens when you get it right
When you do have a clear outcome, the AI ideas get better. More specific. Grounded in an actual problem rather than a capability. The conversation shifts from "here's a cool thing AI can do" to "here's a specific thing we're trying to fix, and here's how AI might help us fix it." That's a different conversation entirely. And a much more useful one.
We haven't fully figured this out at JB Sales. We're still working through how to evaluate ideas consistently as they come in. But asking "what outcome does this serve?" before anything else has already changed the quality of the conversation.
One question. That's the whole thing.
Meghan Brenner is COO at JB Sales and founder of The Operator's Notebook. The Muddy Middle is her weekly newsletter for non-technical operators figuring out AI in real time. Subscribe at newsletter.theoperatorsnotebook.co.
