The idea for what to build next is a guess. It can be an educated guess, based on market research, competitor analysis, customer feedback and so on. In essence, Future work is an option, whose value has some quantity (within a cone of uncertainty), but the means to extract that value is unknown.
Should the team choose to exercise this option, they can then Experiment towards a goal. As part of experimentation, identifying assumptions and seeking to validate - or invalidate - provides better clarification of user needs and aligning future product direction to those needs. We know that Fast feedback improves quality.
A negative result is as useful as a positive result by providing:
steering towards future experiments
a risk-reducing process of discovery
a cost-reducing opportunity, through minimizing wastage from building the wrong thing
By amplifying positives (and even Invalidating an assumption generates learning) a team can still develop a sense of progress, adjusting and course-correcting as necessary, with rapport.
Invalidating an Assumption Generates Learning
This can come from Invalidating an assumption generates learning
How can I Experiment towards a goal in a way that even Invalidating an assumption generates learning?
Should a need not have been met by our work, this too has value as Invalidating an assumption generates learning