What "I need more information" usually means

You're 3 weeks into a decision. You've built a spreadsheet. You've talked to 4 people. You've made a list of risks, a list of options, a list of questions you still need answered.

You are not closer to deciding.

If anything, you're slightly further away. Each new piece of information has opened a new branch, raised a new "yes, but," surfaced a new consideration. The model is getting more sophisticated. The decision isn't.

It's been a while since the last new piece of information actually changed anything. The variables you're weighing now are the variables you were weighing 3 weeks ago. The spreadsheet hasn't told you anything since version 2. The work is generating output. It has stopped generating decisions.

There's a useful piece of research on this.

In 2017, the Journal of Economic Literature published a 40-page review by Russell Golman at Carnegie Mellon, David Hagmann at Harvard, and George Loewenstein, also at Carnegie Mellon. The title was "Information Avoidance." The central finding is unsettling. People will pay real costs (time, money, opportunity) to avoid information that would force them to act in ways they don't want to act. Even when the information is free. Even when it would be useful. Even when there's no strategic reason to look away.

The literature calls this "deliberate ignorance." The deliberateness is the unsettling word.

Sometimes the avoidance is the work itself.

This is hard to see in yourself because the output looks right. The spreadsheet is real. The conversations are real. The risks listed are real. The work has a shape and a vocabulary and the texture of rigour. The only thing missing is the decision.

A few signs worth checking honestly.

The same question keeps appearing in slightly different forms.

You keep "needing one more conversation."

Each new piece of information opens 2 new branches instead of closing one.

You can list the inputs to the decision, but you can't name what it would cost you.

You feel productive while you're working on it, then deflated when you stop.

If 2 or 3 of those are true, the next spreadsheet is unlikely to help. You're working on a different problem now.

Three questions worth sitting with, in order.

What would I do if I had to decide today, with what I already know?

What am I hoping the next piece of information will change?

If new information doesn't change it, what then?

The third question is the one that matters. Most stuck analysis is propped up by a quiet assumption that the next data point will dissolve the problem. It usually won't. The problem usually isn't a data problem.

Good analysts know when more analysis will help and when it's pretending. That sense is part of what rigour actually is.