Minimum viable product

<p>When a company starts, it must define its initial offering. The idea of an <a class="wikilink" href="/mvp/">MVP</a> is to come up with the minimum list of features that would allow the company to explore and validate the market potential of their idea. </p> <p>In the book <strong>Inspired</strong> ([[<a class="litnote tooltip" href="/literature/@cagan2018">@cagan2018<span class="tooltiptext">Inspired: how to create tech products customers love</span></a>]]), Marty Cagan argues that the "P" should stand for <em>prototype</em> instead of <em>product</em>, therefore it should be <a class="wikilink" href="/minimum_viable_prototype/">minimum viable prototype</a>. </p> <p>The argument is that if one tries to build a product at the discovery phase, it will lead to all sorts of contradictions (if it's a product, we should be able to sell it, but it may not be as high-quality as expected, which means engineering teams will take longer to build it, etc.)</p> <p>I believe to compensate for miscommunication, calling it <a class="wikilink" href="/minimally_valuable_prototype/">Minimally Valuable Prototype</a> can be even more accurate. It is a prototype and it should deliver value. </p> <p>Developing an MVP for <a class="wikilink" href="/science-based_startups/">science-based startups</a> is hard, because there's already a specific insight that is going to be commercialized (<a class="wikilink" href="/solutions_in_search_of_a_market/">solutions in search of a market</a>). </p> <p>The crucial task of <a class="wikilink" href="/product_development_cycle/">product development cycle</a> for an MVP is limiting the number of features. That means, defining <strong>what not to build</strong>. (Which is specified in the <a class="wikilink" href="/product_requirements_document/">Product Requirements Document</a>.)</p> <p>The objective of the MVP is manyfold. On the one hand it should allow to test the <a class="wikilink" href="/market_thesis/">market thesis</a>, meaning that there's someone willing to pay for the solution. On the other, it should demonstrate which core features are relevant for different users. </p> <p>From the perspective of a <a class="wikilink" href="/product_manager/">product manager</a>, an MVP also allows to map out <a class="wikilink" href="/who_are_product_stakeholders/">who are product stakeholders</a> in a given context. </p> <p>And the constraints are also important: the MVP should be developable with the resources already available. </p> <p>At this point, I like to bring in <a class="wikilink" href="/jobs_theory/">jobs theory</a> to the discussion, since it allows to map out the metrics the users will use to judge our offer. </p> <p>One of the complicated aspects is that <a class="wikilink" href="/premature_scaling_can_limit_system_iteration/">premature scaling can limit system iteration</a>. Careful consideration is necessary to optimize the learnings that an MVP will produce.</p>