Validation and Technical Risk
You know the stats, according to this 2016 Nature study, more than 70% of researchers have tried and failed to reproduce another scientist's experiments. I'd strongly recommend that you repeat your proof-of-concept experiment, ideally in a different context or with an independent third party. This might sound like a pain, but it's better to test that the fundamentals are sound sooner rather than later.
1
Repeat your proof of concept
Check what evidence you need
Ensure that you're gathering the evidence required to convince the people that matter. This require a conversation with your institution, an investor or buyer to understand what evidence would most convincingly communicate the value of your solution.
Who you could work with
There are multiple ways to validate your proof-of-concept. You can repeat the experiments yourself (not ideal), or work with an independent collaborator. Companies providing support services to enable repetition include Emerald Cloud Lab (established to reproduce research experiments), and Science Exchange (which enables outsourcing research services to independent labs all over the world).
2
Review the risks
Although working out everything that could go wrong might sound pessimistic, this is one of the best ways to pre-emptively take action and ensure that things run smoothly! Download this template framework which provides examples of some potential risks to get you started.