The output is a proposal a funding decision can be made against, and its main quality is that it could reasonably conclude no. A case that can only recommend building is advocacy, and everybody reading it knows.
What the case has to answer
Which market problem this addresses, carried forward from the validated set rather than restated. If the problem cannot be traced to evidence, nothing after it is worth reading.
Who has the problem and how many of them there are, with the method shown.
What we would build and what we would deliberately not build. Scope stated as a boundary rather than as a list of everything anyone asked for.
Why this company, which is the distinctive competency argument. Without it the case describes an opportunity rather than an opportunity for us.
The financial model. Revenue, cost to build, cost to serve, and the shape of the curve over time. The assumptions matter more than the totals, because they are the part that can be checked later.
The risks, quantified. Not a list of things that could go wrong but the two or three that would change the decision, with what it would cost to reduce each one.
What happens if we do nothing. Frequently the strongest section and the most often omitted.
Bottom up market sizing
Build the estimate as arithmetic somebody can argue with. Count the organisations or people in the defined segment. Estimate the proportion with the problem, using the pervasiveness evidence rather than a guess. Apply a reachable share given the channel that exists. Apply a price supported by the willingness to pay evidence. Show each step.
A number produced this way is almost certainly wrong and is usefully wrong, because a reviewer can point at the step they disagree with. A top down number taken from an analyst report can only be accepted or dismissed.
The Product Canvas
A Product Canvas is the usual container. Its value is that it constrains the case to one page and forces the market problem, the target segment, the competitive alternative, the distinctive competency and the financial shape to sit next to each other. Contradictions that survive a twenty page document rarely survive a single page.
Presenting it
The audience is deciding between this and other uses of the same money. Lead with the recommendation and the number, then the evidence, then the risks. Executives who have to extract the ask from the appendix will supply their own conclusion, and it will be the one they arrived with.
Where each syllabus puts it
Pragmatic Institute calls this the Business Plan and places it at the head of the Business category, and the Focus course teaches it as proposing to the business.
Scrum has no artifact for it, because Scrum starts once a product is being built and says nothing about how it came to be funded. A Product Owner still meets the same requirement whenever an organisation decides whether a product is worth starting or continuing, and everything above is what that decision needs.