An MVP is a Minimum Viable Product. To many, an MVP becomes viable once you can sell it.
You should treat your MVP as the beginning of a trial period. That means that at launch the product has enough substance so that you can learn and get helpful feedback during the "trial period".
I would recommend reading Lean Startup by Eric Ries. He says "MVP is the version of a new product that allows you to collect the maximum of validated learning about your end customers, with least effort!"
The most important first step in building an MVP, and interestingly the one that most commonly gets skipped, is talking to your potential customers. Find out what they want and what they are willing to compensate on. Like how I compensate for my height by being, according to my kind friends, "deceptively athletic".
According to CB Insights, one of the major reasons a startup fails is that there is no market need. So, start there.
Maybe you don't need to compromise on quality and are ready to go to market with a MAP (minimum awesome product). If you need inspiration, think about the guy who wrote the Map song for Dora the Explorer. That's a three-sigma success case of an MVP.
What are you building? A product or service that is faster, sleeker, more intuitive than what is out there? Or, are you creating something new? According to the founder of 500 Startups, if you are creating something unique in an industry with zero alternatives, your MVP can actually be your MAP.
Most importantly, make your product easy to use and make sure it solves a problem. Customers, especially potential ones that haven't bought your product yet, will try to pull you in many different directions. Stay focused on your goal, and don't give into every customer suggestion. You don't have to be as closed off as United Airlines' customer service, as it may help to use customer feedback to guide your product development.
It is easy to get distracted by adding new features and design updates. Just think about if those are helping make your product more convenient and the user less frustrated.
So how do you do it? Start by setting a timeline. Establish clear goals and milestones that you will hit. Be as specific as you can and use numbers, such as talking to 10 potential customers before celebrating National Pizza Day on February 9th.
Launching your MVP is the start of a long experiment, so don't give up too easily. If your MVP fails miserably, try targeting a different customer or industry, and focusing on different key features of your business. Be smart about those changes and monitor them as regularly as you monitor your fridge when you're kind of hungry but know there is nothing good in there.
In short, even though both are important, you should focus more on perfecting your understanding of the user and market need than building a beautiful, bug-free product. Oh, and don't forget to think about how you are going to make money from it.
I am a 25 year-old venture capitalist and amateur stand-up comedian living in NYC.