By doing the one feature people care about well. Don't look at huge products, look at really small niche ones. e.g., there are lots of small businesses that ne

By doing the one feature people care about well. Don't look at huge products, look at really small niche ones. e.g., there are lots of small businesses that need to write up bids for work and email the resulting quote as a pdf. That's it! A lot of the competition is stuff like Salesforce or JobBoss or other ERP tools that are gross overkill for the 20-person welding shop, not to mention way outside their budget. Or, as I found out from browsing commercial lawn care forums, people want scheduling apps that can take before and after pictures of lawns to email to the homeowner. There's another fairly small project that's useful and saleable. You'll find that if you go down this path, the biggest challenge is going to be marketing and customer acquisition, not building the product. I'm always tempted to build small apps like this, but I keep going back to embedded work because it's more fun!