How about going backwards? Find the first hundred users or at least 10-ish people who actually has a pain point, and build the product for them.