One of the product was consumer and had a hockey stick growth curve obvious in the first week. So it was obvious to keep going on.
Another product was profitable and we could acquire new customers through advertising. So it was obvious to keep going on.
For the one's that didn't do good, initially I wasted a lot of time doing things which won't meaningfully help with adoption. Some of them are:
1. Building too many features.
2. Building too many platforms at launch.
3. Spending time on PR
4. Figuring out the perfect font for the website.
5. UI polish.
Over time you develop better heuristics. You can't ever be fully right but you just get somewhat better judgement.