What a CFO should hold a roadmap to

Roadmaps fail in a predictable way. They list projects, phases, and a go-live that photographs well. They do not list the baseline, the owner, or the work that will be refused. A CFO is then asked to fund a sequence of activities and hope the benefits arrive as a mood.

Hold the document to four questions. What will be numerically different, and from which baseline? Who inside the company owns that number after the consultants leave? What are we explicitly not doing in the next four quarters? What internal capacity does this consume, by name, not by a generic 'client team'?

Capacity is the question vendors slide past. A roadmap that assumes your best controller can also run design, data, and the existing close is not a plan. It is a wish. Move the date before you move the people into a heroic year.

The useful roadmap is short enough to be argued with. If it cannot be read in one sitting and challenged in the next, it will not govern the program. It will decorate the kickoff.