Things I learned from doing an MVP

By Dmitry Kozachenok

For the past few months, I have been helping my sister build a web service in Russia, designed around providing professional styling advice to females from all over the country and helping them look fabulous no matter what their body type is and how much money or time they have. New market, not proven business model, small budget and a part-time team were the inputs to the formula and here is what I learned as a result.

#1 Ads don’t really work. Particularly in print

My sister works as a stylist at a large fashion magazine in Russia and she was giving an interview in it on what it is like to be a stylist. Naturally, she mentioned the start-up she is launching in the interview and talked about the value proposition. Can you imagine better PR? 2.5 million readership, endorsed and highly targeted advertising and completely for free!

When I was asking classmates to guess how many people visited our site in the first three days after the magazine with the interview was published, their estimates were around 100,000-200,000. The real number was 150 people. What if I had to pay for this form of ad? That would have been a disappointment.

My hunch is that word of mouth is the new form of advertising. The product needs to promote itself my adding great value to its consumers. Don’t spend a penny on advertising unless you have several million in the bank, a product-market fit and are ready to scale.

#2 You still should get press coverage

Knowing that by a certain date people would be able to learn about our product helped us tremendously. We spent eight months thinking about the idea, doing smoke-tests, different landing pages and testing value props on Facebook. But only when we learned that in two months there would be an article about us published, we really started building the MVP. Press coverage motivates you to do build something that works by a certain date with the limited resources you have. That’s when you really start ruthlessly prioritizing product features and when the team with full-time jobs is motivated to pull all-nighters for you. The kind of coverage you are going to get will probably not expose you to too many people, so don’t worry if the product is imperfect, just get it in front of some users.

#3 A/B testing at an early stage doesn't work

You simply don’t have sufficient traffic. Buying it is too expensive. Instead, get qualitative feedback from your friends and family. Even the few users you already have won’t probably respond to your request to talk, so do leverage your network – it is there and it is for free. Get their thoughts on the product itself, as a solution to their problem that it is providing. What do they like and what do they not like about it? Make sure you get the core product right and only then think about the color and size of your buttons.