1. Start with listing markets that have a low degree of competition, but don't have a mega-monopoly owning them. These will mostly be small markets.

2. Examine the problem space within each and see if new technology (SW / HW) can deliver 10x improvements.

3. Determine whether these markets are a short enough hop away from deeper ones.

It's surprisingly hard to get a "map" of existing markets, but am curious about those the community can readily identify.