Even as an experienced builder, I still make rookie mistakes.
My biggest one: jumping straight into the build.
A few weeks ago I hit a wall on a project and got on a call with someone whose advice I trust.
Their advice, more or less: ask more questions. Better ones.
So I did, and three things changed:
- I understood the problem properly before touching anything
- The people I was building for started seeing their own problem through a different lens
- Our calls got more productive, and I left each one with a clear direction
My builds used to take 4 to 5 sessions. Now they take 2: one to build, one to audit.
The second change is the one I didn't see coming, so that's what this post is about.
The questions are for them too
I used to treat questions as my homework. I asked them so I could build the right thing.
What surprised me is how much the questions do for the people answering them.
Most teams have never had a reason to say their process out loud, step by step. It lives in someone's head, a few Slack threads and a spreadsheet tab.
When I ask them to walk me through it, they hear it at the same time I do. Sometimes someone stops halfway and says, "Actually, that's not how it should work."
That's the real start of the build. A system can only be as clear as the process under it.
The four questions I prep with
These are the questions I bring to every call now, and why each one is there.
1. Once this step is done, what has to be true before the next one starts?
Teams usually know their steps. They rarely say what each step has to pass on to the next one. When they answer this, the gaps between steps show up, and the gaps are where work goes missing.
2. When the information turns out to be wrong, who corrects it, and does the original stay?
Every system gets bad data at some point: a wrong date, a duplicate, a name spelled two ways. This question gets the team to decide who fixes it, and whether we keep what it said before. If nobody decides, the system just overwrites it and no one can tell what changed.
3. What evidence shows this is actually working? Who owns it after that?
"Working" means something different to everyone in the room. Asking for evidence turns it into something we can both check. Asking who owns it gives the build a home after I hand it over.
4. Where is that being recorded right now?
It's the simplest of the four, and the one I'd keep if I could only ask one.
That last one once found something nobody was tracking.
It's also the question that most often changes what I build first.
What changes on the call
When the questions are good, the call stops being a status update. People stop describing the tool they think they need and start describing the work.
They leave with a clearer picture of their own process. I leave knowing what to build first, and sometimes what not to build at all.
The full method
I turned the full method into a free 6-page PDF: a 4-step prep, questions for 4 meeting types, a one-page brief, and an AI prompt that drafts it from your notes.
Download the guide (PDF, 6 pages)
It's the most basic step there is. I still need the reminder.
If you're about to start a build and want a second pair of eyes on your questions, I'm happy to talk it through (opens Google Calendar in a new tab).