To begin with, a Minimum Viable Product ( MVP) isn’t a Prototype or a fully-fledged product with all the features developed to perfection.
What is Minimum? To understand Minimum, you need to first define what you are looking to achieve or the hypothesis you are looking to prove. Just focus on the immediate goal. i.e. the next most important thing to do, the one thing you must prove before you take the next step.
One way to define the MVP is to create a list of everything you need to prove your hypothesis. Rank them in priority order with the most important first. Now, re-evaluate each item from the top and ask yourself, do you really need to do it? If not, then remove it or think what is the least you can do for that item. Do this repeatedly until you are left with list of items that you can no longer reduce. This is a good starting point.
Once you have done this, think about what you are left with, is it viable? i.e. Can it be done cost- effectively and would you be able to collect enough evidence to be able to prove your hypothesis. If it’s not viable then add as many items as needed to make it viable. If needed repeat the above two steps as many times as needed.
For example; Dropbox, contrary to belief, the Dropbox MVP wasn’t a piece of software at all, or a wire frame or prototype. It was literally just a video explaining the concept. Drew Houston (CEO of Dropbox) needed to prove that users would use his product before they made it, hence they quickly validated this assumption by producing a video that demonstrated the concept and asked users if they would want to try beta. As a result, they gained seventy thousand users in a single night. This was enough for them to go out and build the full product.
Also, please note that an MVP is not a one-off. It’s a continuous process and each iteration of your product must be an MVP that proves a particular hypothesis for you to make the next decision and so on.