At this Sydney Agile Business Analysts & Product Owners meetup Erwin van der Koogh set about educating and dispelling some myths about the emerging and illusive term — the Minimum Viable Product, or MVP.
So what isn’t an MVP?
- It’s not what can be built before the deadline
- It’s not what is possible for a given budget
- It’s not the first version that is going to be delivered
- It’s not the least you can get away with.
It is a product, whether built or prototyped or simulated that is designed to get a market signal to test. Erwin related the MVP concept back to Kano model theory making the point that an MVP should not be the “minimum” product that you can get away with — as that would only elicit an “indifferent” response. This might sound pithy in a blog post, but if you learn about these concepts MVP does then rise as a term that is about building products customers will love, rather than a production term wholly associated with being lean and agile. Erwin adapted a couple of Eric Reiss quotes to make the point that an MVP should be that slice of end-to-end functionality that you can get away with, that also behaves and performs as users expect. Or to bring the love into it
A minimum viable product is that version of a product which allows a team to collect a maximum amount of validated learning about true love [sic] with the least effort
Some examples further illustrated the point.
- Dropbox was not the first file sharing service but it was the first that allowed synching to a desktop drive. To test the product a video was made that pretended that the product did in fact already exist to test whether anyone would use it and buy it. The Dropbox video MVP attracted 50,000 sign-ups in 48 hours.
- Zappos, the first online show retailer needed to prove that customers would buy shoes online. The team went to a shoe store, bought a shoe, took a photograph which they posted online. Someone did buy it and while they lost money on their first sale they did prove that customers would buy shoes off the internet.
- Intuit, wanting to help fight poverty in India had a hypothesis that fisherman with access to real time price information would be able to better manage and plan their business. This team created paper prototypes that people signed up to. Behind the scenes someone manually entered and texted market prices to the tens of people who signed up, and elsewhere in the simulation a couple of people pretended to be an IVR and quickly iterated scripts before the service was built.
In the discussion I asked whether or not minimum viable experience would be a better term, and the phrase minimum viable experiment was explored. Theory around the experience economy and the Cynefin framework was explained and again Eric Reiss was adapted with a bit of folly …
A start-up is a human institution designed to find true love [sic] under conditions of extreme uncertainty.
And while this theory is a lot to get one’s head around it makes sense. MVP as both product and research approach is designed for operating environments of extreme uncertainty, and for creating products for an experience economy. I.e. product research that can be experienced by customers in real market conditions.
Attendees asked what the difference was between a MVP and a prototype with the answer that while a lot of MVPs are prototypes, not all prototypes are MVPs. This makes sense if an MVP is a product designed to test for a market signal, rather than other research objectives often associated with concept and usability testing.
The talk finished by encouraging Business Analysts to switch their mindset from a “solution finder” to a “problem understander”. Music to the ears of this designer.
Check out all of Erwin van der Koogh’s slides from this talk on Slideshare.