Netflix began at the mailbox. After enough people proved they were willing to wait 2-3 business days for movie deliveries, the demand for instant online streaming basically created itself.

But who’s the real MVP?

Not who, but what – your minimum viable product. This is often considered to be a stripped-down version of a new product idea that has just enough features to gather measurable learning about itself and its potential market. But this classic definition can be somewhat restrictive, as the Harvard Business Review divides MVP strategies into two camps:

Validating MVP vs. Invalidating MVP

Validating MVP

What is often meant when using the generic term “MVP”; bottom-up strategy

  1. Offer a product that’s inferior to your imagined final version.
  2. Use customer discovery to recalibrate the MVP toward the wants and needs of its targets
  3. Slowly scale your product toward completion, adding more expansive features along the way.

Example of a validating MVP: Snail Mail-only Netflix then scaling toward instant online streaming.

HBR, however, warns that, with a validating MVP, “success proves your model but failure is inconclusive.”

In the case of Netflix, let’s say that the initial snail mail-only version of the company had failed.


-The endgame of an instant online streaming service is a bad idea.

-That the snail mail approach itself was off-target.

MAYBE the movie database wasn’t broad enough.

MAYBE the monthly pricing plans didn’t offer a better alternative than just driving to Blockbuster and paying Blockbuster rental fees.

What HBR is getting at is that a validating MVP often “only invalidates one small way of acquiring customers. Therefore, you could waste a lot of time cycling through one strategic failure after another. That’s why sometimes working backwards is better:

Invalidating MVP

Top-down strategy

  1. Offer a better product than what the final version will be.
  2. Personalize product for each customer
  3. Scale backwards to find a sustainable model

Here, “failure means the business model is doomed but success is inconclusive.”

Example of an invalidating MVP“The Concierge MVP” approach is best explained through the founding story of Food on the Table, a meal-planning service and Austin startup.

“Before building anything, the two founders went to their local shop in Austin. They interviewed shoppers until they found one that was interested in their service.

She got a concierge treatment.
The CEO visited her every week. He came with a shopping list and selected recipes, carefully chosen based on (a) her preferences and (b) promotions in the local store. The list was updated on the spot based on her desires and feedback. Most importantly, the CEO would pick a check of $9.95 for this service.

This was no way to get rich.

But each week, they would learn more about what it takes to make their product a success. They kept adding more customers to their weekly visits, until they couldn’t handle the load any more.

Only then they started coding.” –(Scale My Business)

Of course, this service was unsustainable. But it answered a lot of questions about the general wants and needs that would have to be addressed in a large-scale version of a meal-planning service, all within a very short amount of time.

Which approach will work for you?

It depends on your product. Consider snail-mail Netflix vs. Food on the Table. Though movie tastes and food tastes both can be finicky and dependent upon the moment, the essential difference is that we eat more often than we watch movies that arrive by mail.

Beyond that, Netflix users could keep the DVDs as long as they wanted, but a singular purchase of ingredients couldn’t be eaten twice. So the early adopters of Food on the Table guaranteed a weekly demand, while a very small pool of Netflix users would’ve been too sporadic to offer reliable market data.

Are you able to create a better but unsustainable version of your product?

With an invalidating MVP, you’ll be able to get a lot of very specific, very actionable insight about buyer personas, without all of the waiting time in slow, upward scaling. It also avoids a lot of the confusion over what works, what doesn’t work, and why. But not every product can be tested this way.

While a validating MVP can leave you hopping from one failure to the next, sometimes, the groundwork for a previous mishap can be repurposed for a major breakthrough.

CastleBranch, tekMountain’s parent company, actually began as a nanny finder service. But the founders soon realized that, if they did their job well, that particular family in search of a nanny was now off the market for the foreseeable future. They then pivoted their background screening capabilities for nannies into a broader service for employers to vet job applicants. And the rest, boys and girls, is history (to be revealed in a future blog post).

As for your own startup, if you’d like to learn more about which MVP strategies will work best for you, contact tekMountain today.


This blog was produced by the tekMountain Team of Sean AhlumMike PattonRod WhisnerAmanda SipesBill DiNomeAlexa Doran, and with lead writer Zach Cioffi.


Comments are closed.