Most of us has tried that the nearest customer was one self or a manager. But that's a poor case if you wish to cater to an actual industry such as logistics, food, healtcare etc.
Having been in IT for nearly 25 years I am still stunned whenever I see people simply throwing technology at a challenge or idea with slowing down and actually research and validate the idea and hypothesis.
How do you actual go about this before you start carving the wood?
The steps: - Launch a minimum viable product to test market response. This can be as simple as a paper prototype (a concept or conversation starter), a hybrid of existing tech to address a customer pain-point or an alpha version of software development. This may also be seen (from an outsider point of view) as "people simply throwing technology at a challenge". - Iterate and refine the product to better fit the needs. - Optimise for greater gains. Learn from failures. Pivot if needed.
Businesses are ultimately a manner to meet market demands and solve customer needs. It is meant to apply existing research and knowledge to cater to the customers, and not the other way around. If you are interested in research and validate ideas/hypotheses, then you are better off being in a R&D department of a large company that can afford such cost centre for a much longer term profitability.