Saturday, January 14, 2012

Lean Startup : Experiment on Building Failure

Reading The Lean Startup opened my eyes on how to manage a startup.

The definition of startup, in this book, is any human institution designed to create a new product or service under conditions of extrem uncertainty.

One particular quote of this book just opened my eyes : if you cannot fail, you cannot learn.

So I asked, what is failure ? For my previous startup failure was going bankrupt. I failed, I learned, but the problem is : it took us 9 months to see that we were going the wrong path.

I wasted lots of time and money, and minimizing waste is the mantra of Lean thinking.

In scientific methodologies, you prove an assumption (a conjecture) by doing an experiment.
If no experiment can be made to prove that the assumption is wrong then it's religion, belief and politic but it is not science.

If you cannot fail, you cannot learn, what is your assumption ? what experiment can you make to prove you are wrong ? how do you build failure ?

Building the product is not the most important part, building the experiment that will proove without ambiguity if you are in the right path might be way more important.

"Just do it" enterpreneurship, ie creating the product and just see if it works, might be such experiment... BUT, believe me, it's costly with lots of waste, I hope I will find cheaper next time.

Monday, January 9, 2012

Job security and single point of failure

In software development, there is two ways to achieve highly reliable service.

The first solution is to buy the best hardware, buy the best software, and hire the smartest people to have a highly reliable server. We call it centralized architecture.

The second solution is to buy lots of cheap computing units, with cheap software that can do the same things and hire the smartest people. These units are not very reliable, but connected together, it creates highly available service. We call it distributed architecture.

Why choosing one instead of the other ?
First difference : Centralized architecture has an exponential reliability/cost curve, whereas distributed architecture has a linear curve.

Second difference : Given unlimited money centralized architecture is limited by technology, it's not the case for distributed architecture.

What is the big deal with job security ?
Let's change the labels.

When you are an employee, you are limited by the security of your company.
When you are self employee you develop your network and knowledge to cut any dependencies to a single point of failure.

It does not mean that you can't do both and keep the best of the two worlds.

Just keep in mind that if you want real job security, you have to develop something outside your company. That's what I mean by self employee.

Greater job security leads to greater confidence, leads to better deals.

Sunday, January 1, 2012

First bankrupt, integrity leakage

My first company Easydhome bankrupts this week, and happy that it did not affect negatively anybody's personal life.

I can give lots of bullshit reasons that drove us to this outcome.
A customer did not wanted to pay the invoice.
Why ?
We have not been explicit enough and they misunderstood us.
Why are we dead for such simple event ?
Because we had no cash.
Why ?
Because we accepted a contract that should have gave us more money that it made us spend...
Why we lost money on a contract ?
Bad cashflow management.
Why ?
Because we told me everything should be right, no need for cashflow management.
Why I accepted to stop doing it ?
I protected myself against personal financial loss by seeing apocalypse coming. No personal loss No problem I thought.
Why did you think about you ?
I accepted my responsabilities as a president but did not act as a president,
I accepted the consequence of breaking promises, and I let others the power to break them. And it brokes.

I fled conflict and protected myself instead of my company. I followed a path I did not believed.

Some stupid people have power but doesn't want to deal with responsabilities.
I did the reverse, I let the power and was ok to deal with responsabilities.

Now, I know that only two right solutions existed :
  • Saying no for every decisions that I did not firmly believed as right, changing belief as an option but making no compromise
  • Being fired from my president function, and letting it go,
When you have power, your duty is to do the right things, thinking of myself first was definitively not the right thing.