The most expensive assumption is the one nobody questioned.

Every big decision you make sits on a pile of assumptions, and most of them you never say out loud. You assume the customer wants this. You assume the data's clean. You assume the partner is who they say they are. You assume the thing that's always been true is still true. Some of those are rock solid. One of them, usually, is quietly wrong, and it tends to be the one you never thought to check.

The risks you can see are the easy ones, in a way. You plan for those. You put them on the slide, you give them an owner, you pad the timeline. The one that actually wrecks things is the assumption sitting so far under the plan that nobody even sees it as a choice.

I learned this the hard way, years ago, on a business I helped build. The product was good. It solved a real problem, and we built it well. It still went nowhere, because the people we'd tied ourselves to carried reputations in that market that none of us had bothered to look into. We'd assumed they were credible. That assumption was holding up the whole venture, it was wrong, and by the time we understood that, the money and the months were gone. Nobody had thought to investigate the one thing that turned out to matter most.

Here's a happier version of the same lesson. A company I worked with was convinced that growing meant hiring more people to hand-key old records, because everyone knew those records were too messy for a machine to read. That "everyone knows" was the whole game. Nobody had actually tested it in years. When we finally did, it fell apart, and the entire economics of their growth changed in an afternoon.

So before you commit real money or real time, go hunting for the assumption you're not examining. Say the quiet parts out loud. What has to be true for this to work? Which of those things are we sure of, and which are we just used to? If this one belief turned out to be wrong, how much of the plan comes down with it?

That last question is the one worth sitting with. Find the belief holding up the most weight and go poke it while poking is still cheap. Testing an assumption early costs you a conversation or a small experiment. Finding out you were wrong later costs you the whole thing.