Also, Rust is very hot. You will get investor confidence if you tell them your back-end is in Rust because then they'll know it can never crash or have bugs.
In May 2020, the timing will be perfect for some ideas, but not for others.
You're also not likely to succeed if you start with an idea or a problem and then try to find customers. Start with the customers -- people who trust you enough to buy something from you, even though you're just starting out.
That's the real barrier to sales for most startups: no one wants to be your first customer or your guinea pig. They know there's a 99.9%+ chance you'll go out of business.
People rarely take that leap unless they know and like you already.
OK, how about a digital "yarn board" for crazy conspiracy theorists to collaboratively piece together things?
I help non-technical people build companies for a living (2-5 products every year).
Even some people with time, money, and connections can't make it happen. It's incredibly silly to suggest that execution is basically an afterthought.
- niche, saas, privacy first search engine (for developers, lawyers, any profession)
- paid search engine for quality websites only, such as hobby, personal blogs and non for profit in general, also company blogs can match that criteria, i.e. exclude websites with ads (like reddit), popups (medium, quora) and so on
- I feel like we have a place for one more "not evil" git hosting tool without telemetry :)
- email maid easy like all in one written from scratch on high level language like python, ruby, etc.
Maybe I'll do them all when I can sustain myself enough to work fulltime on side projects. Search engines are quite capital intensive though.
Also as many wrote here, you don't have to wait for one year, just start today with general planing, work on the business model, talk with customers, talk couple of agreements in advance to reach later when you'll have a product and so on.
Email right now is extrimely complicated ("s" in smtp is for "simple" btw) and in order to do it yourself you need to understand many parts of it like dmark reports, constant monitoring of delivery, spam, spf, dkim etc.
There are couple of bare-bones opensource solutions to host it yourself but if you do it from scratch you will go either with postfix or exim. They are written in "low-level" languages like C where extremely hard to write proper software without buffer overflows and so on.
I think there is a place for innovation for better security and user experience as well. You can sell additional services like better spam and malware detection or delivery monitoring and support for enterprise.
I wish I can do it myself but I'm already occupied with too many side projects :)
You should probably get real world experience. Problems looking for a solution will come to you much more organically than asking HN.
edit: maybe invent a crappy pointless software removal tool
U WOT M8?