Filter this into a prioritized list that you see cursory potential to solve a problem, ordered by size of the problem (in terms of market size or cost). Take #1, create some assumptions and subject it to a "validation towel off" by speaking to the consumer and the customer (can be one of the same, may be different). Lather, rinse and repeat.