Most software projects do not fail during the build. They fail at the point where someone agreed a price for work nobody had defined yet. The estimate becomes a commitment, the commitment meets reality, and everything after that is a negotiation about who absorbs the difference.
What discovery actually produces
We run a short discovery before quoting anything substantial. It is a fixed, paid engagement with a defined output: a scoped build plan, an architecture outline, and a costed roadmap. It usually takes one to two weeks.
Clients sometimes read this as a delay. In practice it is the opposite. The alternative is not a faster start — it is a start in the wrong direction, discovered in week six, paid for by someone. Discovery is the cheapest week of the project because it is the week where changing your mind costs nothing.
Why it makes projects cheaper
It also changes the conversation. Once the plan exists, the discussion moves from 'how much for a website' to 'which of these things do we actually need first'. That is a much better question, and it is one you can only ask when the options are written down.
If the discovery says the project should not happen, that is a good outcome too. We would rather tell you that in week one than bill you for four months of building something that was never going to work.