Hacker News new | past | comments | ask | show | jobs | submit login
A new approach to minimum viable product (humbledmba.com)
48 points by benehmke on Nov 22, 2011 | hide | favorite | 19 comments

MVP is one of the many words which has undergone corruption and softened in meaning through people's personal interpretations (Pivot, Startup...other increasingly vague terms).

The approach put forward here is not exactly new - if you read Ash Maurya's Running Lean (http://www.runningleanhq.com) he talks extensively on using a rich landing page (rather than a 'viral' launchrock-style page) to accelerate validated learning. The classic DropBox video example is another piece of classic Lean case material. What these both somewhat lack is the 'P' of MVP. Product.

I'd suggest this is 'practical application of Lean Startup principles' rather than MVP - people tend to talk about 'building their MVP' and that process hasn't begun in this case. Validated learning is critical, and getting a headstart on it in this way is a useful technique, but blurring the distinction of early-stage customer development and the actual MVP doesn't help the discussion. It makes the conversation around desirable qualities of a public MVP somewhat harder.

A minimum viable product is something that customers can actually use. Something that you can sell, that people are willing to pay money for.


> We did everything possible to not show that we hadn't even started the back-end yet.

Is testing marketing and UI design. Which is great. But it's not an MVP.

The term "MVP" has expanded to cover things that are not strictly products. E.g., Eric Ries refers to a demo video as the "Dropbox MVP" because of the way Dropbox used it successfully to generate demand: http://techcrunch.com/2011/10/19/dropbox-minimal-viable-prod...

Rather than reject techniques like this because they're not technically products, I think we should expand the concept of MVP to include things like this because they serve the same purpose: learning about your market, reducing risk, validating ideas.

I think this can still be considered an intermediate MVP that can be tested. The goal of lean is to eliminate waste, so there is a risk that an MVP with working code has no demand whatsoever.

One drawback to this approach is that it may give you false negatives. What if you landing page copy doesn't work or your AdWords copy is wrong? As with all lean/custdev hacks ymmv. Bottom line is you need to find customers to talk to, otherwise everything is an assumption.

Some other articles that touch on this approach:




There is no "intermediate" MVP, it's a contradiction in terms. It's either minimum AND viable, or it's not.

i don't disagree that there's value in using this stuff. a lot of value. but doing only this is not an MVP. "P" = product, something you're selling, something people are using.

there should be another name for this. MVI maybe? minimum viable idea?

While talking with Bob Walsh on the Startup Success podcast a couple weeks ago, he came up with "Minimal Viable Experience" which I think applies well here: http://startupsuccesspodcast.com/2011/11/show-124-corey-maas...

Wikipedia's article on mvp (I know, I know), includes this:

>The canonical MVP strategy for a web application is to create a mock website for the product and purchase online advertising to direct traffic to the site. The mock website may consist of a marketing landing page with a link for more information or purchase. The link is not connected to a purchasing system, instead clicks are recorded and measure customer interest.

from: [wikipedia](http://en.wikipedia.org/wiki/Minimum_viable_product#Techniqu...)

That seems to directly contradict the implication of 'mvp' that you're using, which is that it MUST be a viable product that customers can use.

EDIT: The best part of markdown is when it isn't supported, the code is very readable. I say leave it!

the wikipedia article seems to me to be referring to an MVP more as a process, and the smoke testing by creating mock websites as one part of that process.

i can agree with that, on some level, but i still feel that to have a complete MVP (process), you necessarily have to have a P at the end of it.

Not even that in my book. It's hoodwinking that could potentially hurt existing businesses as well as creating an atmosphere of distrust if practiced widely.

Nor a new approach.

This isn't MVP (there's no product!), just the beginnings of customer development, no?

The cicles do not make sense to me. Minumum viable means the result of the lowest effort/functionality that leads to a viable product. Using overlapping circles just does not work at all, I'd say

Using minimal without viable is somewhat strange. If you look at a program: What is minimal? Hello World? No, I can still do "int main(){ }"! But wait, this still does a context switch... IMO it just does not make sense to say something is "minimal" (but not viable). it's better to say there is viable and not viable and within viable thing, you can head for minimal, too.

There is a minimum circle and a viable circle so it would follow that they must both contain products.

So the circle on the left contains minimum (crappy as they indicate) products and the circle on the right contains products that are all viable - some overdone perhaps.

I'm not sure why it wouldn't make sense that something is minimal but not viable. Any product that can't sell because it is too minimal would fit there.

And the intersection that they are showing is in fact what you are describing: viable and within viable, also minimal. In other words, the most minimal of the viable.

If you feel a swelling sense of pride every time you publish a mockup or do an a/b test you might be missing the point of MVP -- solving a problem without getting caught up in process.

I got a new term for this. A fake MVP. Think it's a cool concept, as long as you only fool a couple people (around 10 ~ 50 and not thousands) with it to get some feedback.

I am missing the new approach. Maybe this is an explanation of how you applied lean and MVP principles?

Good customer development is filming the trailer and then deciding whether to make the movie.

FWIW: Tarantino & Rodriguez's movie _Grindhouse_ included several fake movie trailers. They were so popular that _Machete_ and _Hobo_With_A_Shotgun_ were turned into real movies with good boxoffice returns; more such projects are underway.

This is not a new approach. Vaporware and other deceptive practices have been plaguing our industry for decades.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact