Your greatest risk management tool is research.

Good quality market- or UX research can't tell you for definitely what is true (such as "customers will buy this product"), but they can tell you what is definitely not true (such as "customers will not buy this product because they are not using the platform for which it is designed").

Of course, caveat caveat, you might get strange things happen like a resurgence in the popularity of cassette tapes, but not on a scale that would support a decent-sized business venture.

Research can tell you what (potential) customers value and care about, what they prioritise, what they struggle with, their habits and attitudes, and the context in which they live and work.

Or, to put it another way, the first thing I do with any new team is to flip their strategy to "what would have to be true?" for each point, circle that ones that fall under my remit (for instance, "does it work?" is more of an engineering question, not mine) and then get to work:

"You should adjust your strategy to sell driving tests to toddlers because it is illegal in that country for toddlers to drive."