I don't have statistics, but let's say a lot fail at step 1 or 2. Generally developers build something without validating it is worth building first. Plenty of great ideas out there, but find a market before you build. If you find the market, you can then figure out monetization and decide if it is worth building, but usually, I'd say the vast majority fail between step 1 & 2 as you have listed them.
Hardest part depends on who you are and what is happening. e.g. if you are a dev who is afraid to get in front of people then usually that is the hardest part and where you will fail.
Gaining market share is about how you advertise and which niche you target first. Within all markets there exists niches you can exploit first, and then expand. The key is finding them and exploiting them. And there is a definite need to spend money on marketing and business development which is foreign to a lot of engineering mindsets at first.
BTW -- not bashing on engineers/developers it is just where most people fail that come from engineering backgrounds IMO.