Go over every requirement and decision and defer or remove any that an MVP won’t require. For example, an MVP car (which took more than a couple of weeks to build) needs an engine and wheels. It does not need doors, a windscreen, or bike rack.
If your MVP has a good foundation and communicates the business proposition well people interested in it can fill in the blanks and imagine the next iteration.
I am building a what I can only describe as "Supplement discovery based on science and not marketing". It requires that I create a database of supplements available on the market, then allow to compare them at a ingredient level to understand which is the best one to buy.
I started building this out of frustration of trying to figure out which supplement I should buy myself. It took a ton of time to build a comparison table ( Here is one example https://docs.google.com/spreadsheets/d/1vQcLEAPk5lcg-4YR_dwD... ).
So I validated with a spreadsheet that people find it useful. And then I started building the data model for representing this across all supplements. And this is where things have gone a bit south. Somehow now I have most ingredient data at a molecular level. I tell myself that I needed to go this deep to ensure data quality (def don't want to accidentally misrepresent ingredients).
I am over this hump now, i.e. I have indexed and categorized every ingredient of over 10k+ supplements. Now I am just building the UI to compare them, but it feels like dug myself into a whole for a good part of the week. Will see what happens once I make this public; if people use it would be hugely rewarding.
But going to the car analogy... my mind is telling me that doors, a windscreen, or bike rack are all nice to have things that translate to various compromises in the UI. Just keeping that extremely simple. But that data integrity is not something I can compromise on (it is like letting car for a ride with an engine that may break).
Don't know how this way of looking at things is true for others. Sometimes I feel like I just have an OCD specific to designing the underlying data architecture.
Maybe the car analogy doesn’t apply. But consider Henry Ford went to market with just one model and color of car.