Start writing a code or creating a design only if you do not know another way to find out if your product is needed.
The minimum viable product (MVP) is a product with just enough features to satisfy early customers, and to provide feedback for future development. There are many arguments around this concept, but the idea is quite obvious: understand as early as possible that your product is needed. The definition of this strategy speaks for “the minimum set of features required to run and test the product.” It also has to be finished faster and with minimal costs involved. But what would you do if nobody needs this product with the set of features you implemented?
That’s why MVP is not about the product, but rather about a strategy. A strategy to avoid building products that people do not need. You can test key assumptions about the product without writing a line of code. Even without a wireframe. A product must solve a specific problem and be useful first and foremost before design is even considered. That’s where a proper research comes into play.
But if you still built a product that nobody needs – forget it and let it go. Eventually, MVP can be somewhere around “unfortunately, it did not work out, but fortunately, I did not spend a lot of resources and time”.