Michael J. Sikorsky | Reports Greatly Exaggerated

The size and scope of Eric’s Startup #1 failure ($40M into /dev/null) gave him the necessary freedom to challenge his prior (but faulty) core assumptions to create Startup 2 (a massive success).

Please make sure to check out Slide 33 where Eric talks about “proportional investment”.

Eric Ries Lean Startup Presentation For Web 2.0 Expo April 1 2009 A Disciplined Approach To Imagining, Designing, And Building New Products

View more presentations from Eric Ries.

The biggest risk for your startup is building something people DON’T want. So, taking the inverse: your startup has a straightforward task: build something people want.

If you build something people want, you win [1]. If you don’t, you lose. And, the losing hurts in ways you didn’t know could hurt [2].

Q: How do you know what people want?
A: You don’t.

This is why so many startups fail: you can’t predict what people will want. [3]

You will have hunches, insights, passions, and a starting point but in the end your success will likely look much different than what you thought.

Flickr was supposed to be a online game.

PayPal was supposed to sell cryptography libraries.

(Read F@W for many more similar examples.)

Don Keough (former CEO of Coca Cola) spoke to why he sold Paramount Pictures. The net of it: he didn’t understand how to build a business around movies. It was too “hit” driven. Meaning, for reasons beyond their control (genius), some movies would do amazingly well, and some wouldn’t. Compare this to selling cans of Coke (process).

Startups can feel like the movie business. It can seem hits driven (a “yuppie bingo” model). The key point is that they don’t have to be. You don’t need to take so much risk with your time, your reputation and your investors capital. [4]

How you ask?

Deploy. Learn. Repeat.

If you can keep iterating and don’t run out of money (or morale) you win.

Movies can’t do this.

As Eric points out in his presentation, eventually IMVU was able to release their software 50 times per day (continuous deployment). Think about the power of that: 50 times per day they can learn about what people want.

I liken this to flying. You can fly a plane a night using the instruments (IFR). But, if those instruments were only updated once per hour, you’d be dead. It is the same with startups: the faster you can refresh your instrument panel, the faster you can iterate, the more likely you’ll build something people want, and end up giving speeches at Web2.0

The deploy, learn, repeat is a PROCESS. It isn’t about being a visionary, genius founder. PROCESS over GENIUS is how you significantly increase your chances of building something people want.

I think Eric is going to have (is having) a transformative effect on the startup/hacker community. But, I wanted to point out Mary Poppendieck. I was lucky to spend time with Mary and take her course on ‘Lean Software Development’ a few years ago and I really credit her with championing on how lean manufacturing principles can be applied to software. Lastly, if this is all resonating with you, you’ll want to buy 4 Steps to Epiphany by Steven Gary Blank.

[1] Once you’ve built something people want, you’re called a company Once you’re a company you’ll call this stuff ‘innovation’.

[2] I know I can’t talk about this… yet.

[3] This is why all those other proxies for success are used by investors. Team, etc.

[4] Yes, I’m aware of the irony.

** You can follow Eric on twitter here: http://twitter.com/ericries (I’m here).

** Audio is thanks to the guys at VentureHacks.