Minimum Viable Product or MVP is one of the most misunderstood terms from the concept of Lean Startup in my humble opinion. Some people start associating the term ‘viable’ with ‘usable’ for the customer. I have seen arguments about how MVP for enterprise products can’t be too minimum or else the customer wont let you in the next time or overloading of features or users and not buyers etc etc. Sorry, but I think this is a mistake in understanding the concepts of MVP clearly.
MVP is not always something you ship to your customer, its not the output of SCRUM (agile development methodology) that has to be shippable increment and no the customer or user does not have to use it for their business to give you feedback about what to build next, thats not the purpose of MVP but of Agile development.
So what is the purpose of MVP then? lets first understand the key hypothesis behind Lean Startup thinking. Startups as we all know are risky business, a lot of assumptions about the product is in the mind of the entrepreneur from his world view, which most of the time is little to very different from the customer point of view. There are two key hypothesis to be proven before trying to scale a startup:
- Value Hypothesis
- Growth Hypothesis
Value is the real solution to a burning problem for which the customer is actively seeking a resolution and is willing to pay to get your solution (though sometimes the solution might be free and making money from another segment). I would think we are not talking incremental improvement over a bunch of existing product, but something that can be considered as an unresolved issue (Blue Ocean rather than a Red Ocean) for a interested segment.
After we have established that the proposed solution has Value for a segment, then comes the question of whats the growth potential for the solution in the market, locally, regionally and globally. Even though the problem and solution might be real, the potential buyers might be too small to make a profitable and viable business, or the effort and knowledge required might be such that it would be hard to scale beyond certain market or cost can be a major hindrance. Even if the solution has value and there is a market, there is still the question of how to reach out and sell it in a way the buyers are willing to buy it.
All the above are RISK that a startup has to face in its initial phase, if the venture got to see success, there has to be a systematic way to de-risk the venture without spending too much in the wrong things and running out of money. This process is what Steve Blank calls ‘customer development’ and approaching a startup with all this in mind in a scientific manner is the whole philosophy of Lean Startup.
So now the question is how do you go about de-risking the venture, the answer is through series of experiments to test the various assumptions and key concepts in the solution. MVP is a tool to run these experiments, not the only tool, but probably the most used tool. Smoke Testing, A/B testing and Statistical Hypothesis Testing are some of the methods for experimentation to confirm with customer, the theories for the startup solution as well the problem perspective.
MVP can be a smoke test website that tries to establish demand for a solution in the market, or the segment that needs it most. Smoke test does not intend to be dishonest with the customer, instead it paints a picture of a solution to come and get interested parties pre-register for invite or information when it is ready.
At various stage of the product lifecycle the MVP can be very different in its form, but the purpose in the world of Lean Startup is to use it to proof or disprove another hypothesis and thus lower another risk to the venture.
Its probably also worthwhile to remember not every kind of customer or user are ideal target for MVP or experimentations, early adaptors are usually the ones we should be looking for, they might not form the bulk of the user segment when the product flys, but they definitely are those who would be ready to touch your buggy barely working concepts and give you inputs that bring your product in the track to be a reality. Will discuss more on Early Adapters in another post one of these days, thanks for reading.