A MVP (Minimum Viable Product) is a startup philosophy and concept popularized by Eric Ries now widely accepted in the startup community. It’s a way to quickly build a product with the only the features needed and quickly validate it with early adopters to see if your product actually delivers the UVP.
A MVP will include and only include must-have features to solve the #1 problem your product promises to solve. The nice-to-have's and don't-need features are postponed till you've validated that your MVP is a viable solution.
In other words, you must understand exactly what your customers want, and delivery just that. Nothing more, nothing less.
So you want your customers to come, and get exactly what they want so that they will come back. Retention is key, not only because we want them to come back, because we know that we have created a product or service that delivers our solution.
Example: Say you are going to open a coffee shop:
Good MVP: You have a menu with only a few selection of coffee, but all of them taste great. Your shop is clean and tidy with white painted walls and decent looking coffee tables with comfortable seats. Your cashier is polite and you take Visa and Mastercard. No American Express though (we're cool with that).
Bad MVP: You have a menu with a few selection of coffee, but most of them taste horrible. Your shop is below par and your seats are uncomfortable. Also, your store only takes cash and your cashiers are flirting with each other.
Horrible MVP: You have 40 items on your menu - Coffee, frappuccino, espresso, cakes, croissants, biscuits, tea, fried rise, chicken nuggets, etc (you get the idea). You gold plate your store front and your store interior is pimped out with the most expensive furniture and coffee mugs. You put Samsung LCD TVs at every table and the chairs are massage chairs. You take Visa, Mastercard, American Express, and Home Depot gift cards.
We've already asked you to build a service-first product in the last section, but just in case you skipped it, we'd like to mention it here again because we feel this is very important.
Before you go ahead and put yourself in the "build" mentality, I'd like you to slow down and think about whether or not you can deliver your solution without a product (What I'd like to call the 1st version or smaller version of your MVP). This is the easiest way to start your business, and having a business that already makes money already makes your solution valid, even through an online presence.
For instance, if you want to build AirBnB, a marketplace for unused rooms, then what you could do first is not build a whole fully functioning marketplace, but think about something equivalent but smaller.
What you could do is find 10 friends with unused space and are willing to rent that space out. Then you can find travel agencies or put up a simple Wordpress website with the availability and price information on the properties.
What is happening here is that you would have delivered your solution on a smaller scale, made some money, and have validated your solution. At this point, it'd be very easy to move forward with more confidence to build the full "automated" version, or to get funded to build it.
Once you have a business that is making money, there will just be a lot more confidence to invest more into it, whether by you, by a technical co-founder, or by investors.