On the shortcomings of “Minimum Viable Product”

Christina Wodtke recently wrote “Getting the V Right”, addressing a common failing of Lean Startup practice — successfully establishing viability in your Minimum Viable Product (MVP). I commented there, but felt it worth expanding here.

The MVP Incantation

My frustration with MVP comes from its reckless use in product management. When launching a feature, I’d hear about “We just need to get our MVP out.” But never was there any attempt at determining viability. What product managers actually meant was, “next release,” but used “MVP” to suggest savvy and greater likelhood to succeed. Christina attempts to address this by coaching people on how to appropriately articulate viability, even resorting to the grade-school-essay canard of the dictionary definition. The problem is, most folks who are misusing MVP are already a lost cause — they’re cargo cultists hoping an incantation gets results, and no amount of guidance will change that core behavior.

Viability is unpredictable

MVP rests on an assumption that you can pre-assess something’s viability with reasonable confidence. However, viability can only be understood in retrospect — you can try to predict it, but really you won’t know until it’s out there. I suspect this is why so many people punt on defining it, in favor of “just get something out and see how people react.” But then this just turns into a resource-wasting exercise in throwing spaghetti, hoping that something sticks.

“MVP” doesn’t galvanize and inspire

However, even if we believe that MVP is an appropriate tool, the confusion in its use suggests a different issue — that while it’s proven catchy enough to spread, it’s nebulousness and abstractness limit its utility. Nebulousness leads to too much variance in how it can be interpreted, and abstractness means it doesn’t galvanize a product team. No one gets excited about launching an MVP. It lacks punch. I much prefer models such as Brandon Schauer’s Cake Model of Product Strategy or Spotify’s vehicular one:

These models communicate what’s most important — that at every stage, even the very first, you must deliver something that feels complete, and delivers useful functionality and/or delight. I’ve personally seen teams shift from MVP to “cupcake,” and, in doing so, shift focus on delivering some bare minimum to something that they can get enthused about.

Maybe we’ve been using the wrong kind of viable?

I teased about the dictionary definition, but there’s actually something in there that’s valuable that I’ve never heard discussed in the context of MVP:

(of a seed or spore) Able to germinate.

We’ve been so focused on economic viability, that we overlooked the origin of the word “viable”, rooted in the word “life.” The common thinking for MVP is “what is the least I can do to deliver a product that doesn’t fail”, but wouldn’t it be more interesting, and inspirational, if we thought, “what can I deliver that could take on a life of its own?”

One thought on “On the shortcomings of “Minimum Viable Product”

  1. The cake model isn’t an apt analogy, because the three steps required to make a cake are also required to make a cupcake. I get what its saying, but i just found that annoying.