Most tech founders start their journey by writing a lot of code. However, many times it’s better to begin with validating product need through a quick non-code/ light-code exercise. E.g. if you’re planning a market-place, you might try to get a couple of transactions done by matchmaking supply-demand simply over email/ offline. Or you might set up a single landing page to collect emails and try digital ads to figure out cost of customer discovery.
Protecting the idea, big bang launch, stealth mode, wowing customer at first experience, etc are all fine ideas if you’re very sure of who your customers are and what they want. But if the best response to “How do you know this is needed?” is on the lines of “We did a survey” or “Our friends told us/ We think so”, then you’ll be much better off doing this first, unless you’re the domain expert. If possible, do it before even quitting your job. It’ll give a lot of real sense on the problem at hand & customer view.
Writing code shouldn’t keep you waiting to discover & work with your customers.