Stop Building What Nobody Asked For: What 7 Discovery Calls Changed

Seven conversations changed the plan

Over a couple of weeks, I did seven discovery calls. Real conversations with people I was building for: founders trying to get out of their own way and professionals trying to become the person on their team who actually knows how to use AI.

I went in thinking I already knew the answer.

Then I did the harder thing. I listened.

Seven conversations. The same fears, desires, and questions appeared again and again. When something repeats like that, it becomes evidence.

The calls changed what I thought was worth building and teaching. They reminded me that nothing useful gets decided in a vacuum.

The screen makes a room look small

Say 300 people see something you publish, and seventeen respond. On a screen, that can look small.

Now picture those 300 people in one room. That's a packed workshop. Seventeen people cared enough to walk up and tell you what landed.

The screen makes a room look small. It isn't.

That shift in perspective matters because building in public can turn people into numbers. Discovery calls reverse that. They put a voice, a situation, and an exact sentence back behind the analytics.

You learn things a dashboard can't tell you: what someone is afraid will happen, what they already tried, where the process breaks, and what “useful” would mean in their actual day.

Nothing real gets built in a vacuum

The first mistake is deciding alone.

You decide what customers, an audience, or leadership needs, then build it in a room by yourself. A founder designs an offer nobody asked for. A professional spends a year improving the thing leadership never rewards.

The fix costs nothing. Have real conversations with the people you serve. Ask what they're afraid of, what they want, and what they've already tried. Listen for repetition.

One person's opinion is a data point. The same need expressed five different ways may be a pattern worth building for.

Then give their exact words to your AI beside the thing you're considering. Ask it to organize the responses, separate repetition from one-off comments, and identify questions you still need to answer.

AI can help you see patterns you're too close to see. It cannot replace the conversation.

The second mistake is executing alone.

AI can let one person start and operate more than ever before. But look closely at almost any “one-person company” and you'll find a network: licensed professionals, contractors, partners, platforms, vendors, customers, and other businesses carrying work the founder doesn't.

One person on paper can still depend on a whole web of people.

You're not on an island, especially if you want free time and a life.

AI changes how much manual work the team has to carry. It doesn't eliminate the need for people, relationships, or accountability. A small AI-native crew and the right partners can move like a speedboat instead of a cruise ship.

Decide with the people you serve. Build with people rowing in your direction.

One thing to do today: ask for useful feedback

Send this to three people close to the work you're building or improving:

“I'm improving [the offer, workshop, process, or work product]. You don't need to be nice about it. What is the first thing that feels unclear, incomplete, or harder than it should be? What would make this genuinely more useful to you?”

Don't explain or defend the work when they answer. Save their exact words. Give the responses and the current version to your AI. Ask it to identify repeated patterns, separate evidence from inference, and suggest what could change.

You still decide which change belongs.

If this is your kind of conversation, join me for the next Seed & Society® newsletter.