Lean startup no funciona en la mayoría de los starups del SE asiático

“Lean” doesn’t work for the majority of startups in SE Asia

Guyi Shen 

The current dominant meme in the startup community is the “Lean” startup.   Take a look at the 52 exhibitors on Echelon 2011′s site , count how many of the startups are either a  social network, a marketplace or  a platform?  The majority of the startups all are primarily relying on network effects as a major part of their product value.  The current “Lean” approach doesn’t work for network effects startups.  Lean is a method of testing the assumptions and hypothesis of a business idea in an iterative manner whilst validating your product/market fit before you ship a complete product.   The popularity of the method stems from the promise that it derisks the product development/marketing so that significant funding is raised AFTER product/market fit in order to scale the business.

Lean as a philosophy is great, however in practise it is severely lacking right now with respect to network effects startups.   Mostly because the user acquisition phase(i.e. scaling) has always been seen as something you do after you get product/market fit.   My opinion is that scaling is something you must do before you reach product/market fit.

Eric Ries  coined the term “Lean”, but the founding father of Lean is Steve Blank  and the lean manifesto is his book Four Steps to Epiphany .  The philosophy of Lean can be described as using the scientific method of hypothesis generation(a creative process), and experimentation(a logical process) to invalidate assumptions.   The vast majority of  “Lean” practises to product development were designed by entrepreneurs working on software that doesn’t rely on network effects.   Current thought leaders in “Lean” methodology include Steve Blank , Ash Maurya , 37 Signals and Eric Ries .   However the majority of current discussion explicitly ignores the unique challenges of  products that rely on network effects.

Because of this fundamental distinction between network effects and non-network effects startups, there is now a tendency to apply the wisdom that worked so well for enterprise software to network effects consumer startups, which does not translate.   Ironically, one of the thought leaders in this space, Eric Ries himself was CTO of a network effects startupIMVU .

More specifically, let’s take the essence of “Lean”, customer development.  The premise is that by questioning and testing the underlying assumptions of your customer’s problems, your solutions, your business model, you are much more likely to build a product that your users will use.   Fantastic premise, however all the case studies, processes, practises currently en vogue in the “Lean” community are based around enterprise/consumer products without strong network effects.   The fundamental difference is that for a network effects startup, a large part(maybe even the majority) of the value to the user is based around network effects.   Therefore THE  core assumptions of your product value can only be tested with a critical mass of users using the product.

At a fundamental level, the value to a user of a network effects startup has a multiplier effect given a critical mass of users, the feature set in isolation does not get to a minimum activation level(how useful would Facebook be if you were the only person on there?).   Given this HUGE caveat, any sort of “Lean” startup approach for network effects startups must take this into account, we must first solve the “lean scaling” issue first.

We need users before we have a viable product.   Sean Ellis first spoke about this at Lean Startup L.A .  Instead of finding product market fit and then scaling, we need to scale first and then find product market fit.

The poster child for this is paypal, they spent tens of millions of dollars  in its first year in order to acquire users, before finding product market fit.  Airbnb spammed the hell out of craigslist  in order to go to critical mass on one side of the market.  Myspace spent a ton on emails, ads and parties.  Youtube gave away a boatload of prizes  to build up their video uploaders.   The lone exception, Facebook, other than some minor levelblackhatting (scrapping data, spamming emails) grew without spending a ton of money on user acquisition.

So what to do when you are not Facebook, nor have money for user acquisition before product/market fit, nor even a ready made userbase  of people willing to try your product.

Welcome to my problem of trying to build a network effects startup from Singapore.   Before you can apply lean, you need to solve a more fundamental problem, scaling with no money and shaky product market/fit, or as I like to call it, “Lean Scaling”.  How do you scale(for free or very cheap) in a way that lets you actually test your assumptions?

Some stuff we are experimenting with;

1.  Segmenting as narrowly as possible.   Theres no way around user acquisition and sometimes you just have to spend the money, but by segmenting tightly, you can control your user acquisition costs to a very tightly targetted segment.   Segmenting geographically means we are targetting only Singapore, and segmenting demographically means we target a particular user profile.   A double bonus for tight segmentation is that if you manage to “burn out” a particular segment through invalidated assumption testing, you can start fresh with a new segment, asuming of course that you still have the resources to keep things “Lean”.

2.  Building two briges at once, and hopefully they meet at the middle.  In our case I wrote extensively about how we tried to simulate one side of a network in order to test some assumptions on the other side of the market.  You can read about it in detail here.

3.  Generate buzz.  User acquisition is expensive, you can make it cheaper by generating buzz, for example I can do a guest post and in the middle of the post ask for signups for our iPhone app beta here .  This topic deserves a book in and of itself.  We have experimented with generating micro-buzz  in very tight segments.

I wholeheartedly agree with the lean approach to building product, however the majority of discussion right now are centred on case studies from products that do not require strong network effects.   And on a tactical level, this just does not work.

Inside Startup

Posted in: Asia,emprendedorismo,lean startup