Five questions decide it. Here is what answers each one.
Whether you’re taking a product into a new market or going for more share in one you’re already in, the same questions decide it. Some have a sharp tool already; where one doesn’t, the question still gets asked and what you assumed instead gets written down.
“The customer said they'd buy it.”
The needs, ranked by how badly they're served
each one carrying its own evidence — or standing marked as unproven
“It's a big market — billions.”
A size you can defend, not the biggest one you found
several methods reconciled against each other, every figure traceable to its source
“We're better than what's out there.”
What it's worth to the customer, in money
quantified against the alternative they'd otherwise buy, and turned into a price
“Operations say they can handle it.”
Capacity and qualification checked, or the assumption written down
so the thing you took on faith is a claim someone has to retire, not a gap
“We'll go direct and see how it lands.”
Entry options costed against each other
with the scale trigger and the kill trigger set before the data arrives
Same questions either way. The only difference is whether the answer was earned.
What answers each one
Answered as checks rather than a rating. Taking an existing product somewhere new collapses this to capacity and qualification, and the route writes down exactly what that assumed: “Accessible capacity covers the year-3 ramp.” A new product asks more of it — development effort, production readiness, what has to be built before anything ships — and how far that gets specified here is an open question. Where the capacity or the build has to be paid for, this is also an investment decision, and it is weighed against the growth plan it is meant to serve rather than on its own payback. Either way this one is never skipped: it is answered, or recorded as a claim someone has to retire.
And one that runs afterwards
This one isn’t part of the go/no-go. It starts once you’ve moved, and it checks whether the answers above held — actuals against the plan, and the triggers you set before the data arrived. It recurs rather than completes.
Runs on what the other five already recorded, rather than on a tool of its own. This is the question that outlives the decision — actuals against the plan, and the triggers you set before the data arrived. It recurs rather than completes, which is why it’s a review, not a gate.
Easier to follow when it’s one real decision.
The same questions, worked through on a single move — including the one with no tool, which is the one that catches the problem.