Eric Ries @ Business of Software. The Science of Lean Startups

“The Lean Startup: A Scientific Approach to Innovation.”

Most software projects fail. Most startups fail. Most new products are never used. But it doesn’t have to be that way. The Lean Startup is a disciplined approach to imagining, designing, and building new products. By testing our assumptions earlier, faster, and with more rigor, we can stop wasting people’s time.

Bio – Eric Ries is an author, speaker and consultant at the Lean Startup. He blogs on StartupLessonsLearned.com on issues concerning startups and entrepreneurship. Previously, he served as Venture Advisor at Kleiner Perkins Caufield & Byers and co-founded IMVU. In 2007, BusinessWeek named Ries one of the Best Young Entrepreneurs of Tech.

Eric Ries starts with the comment that receiving an invitation to speak at BOS from Joel Spolsky was a bit like most kids getting an invitation from Lady Ga Ga to perform on stage. Makes a plea to turn phones on, get online.

“If you are not online you are not alive.”

Most startups fail and most produce stuff that no one uses.

Web 2.0 startups from 2006 failed within three years.

Green circles mean ‘sold’, pink crosses mean ‘dead’.

Who can we blame?

Just over 100 years ago, most work was done by skilled craftsmen. Frederick Winslow Taylor brought science to the workplace.

“In the past, the man was first. In the future, the system will be first.” FWT, 1911.

This paradigm shift in thinking led to the Traditional Waterfall Product Development where the unit of progress is to Advance to the next stage. This works well where the problem is known, and the solution is known. But it is not appropriate in today’s world.

Today, startups and entrepreneurs live in a different world. A startup is a human institution designed to deliver a new product or service under conditions of extreme uncertainty. Therefore, a startup is an experiment and entrepreneurship is management and you can apply scientific principles to create a new management paradigm tailored to conditions of extreme uncertainty.

The Pivot What do successful startups have in common? They proceed down a zig-zaggy path using what you have learnt on the way. You can increase the number of pivots you have by raising more money or by making stuff happen faster by being more agile.

Agile product development has a different unit of progress, a line of working code. This has been the case in most businesses since WWII.

In lean startups, the rules change again. You need to ask what activities are value-creating or wasteful? In traditional business, value is created by getting products or services to customers. In a startup today the product and customer are unknowns.

The unit of progress in a lean startup is validated learning. in the world of the startup, typically the problem is unknown and the solution is unknown. Lean startups are about learning what we need to learn at the lowest cost. Do you need to develop software to learn that customers don’t want it? Often not!

Minimise the total time and money it takes to get through iterations of the feedback loop, continuous deployment.

LEARN faster

• Customer development
• Five Whys

BUILD faster

• Continuous Deployment

MEASURE faster

• Rapid split Testing

LEARN

Five Whys Technique for continuous improvement. Ask why something unexpected happens – five times… Behind every technical problem is usually a human problem. Make proportional investments in prevention at all five levels of the hierarchy. (Do the minimum possible to fix). Do not do zero, do not do perfect solution that will fix everyting for ever). Fix the cause, not the symptom.

BUILD

Continuous Deployment Principles

• Have every problem once
• Stop the line when anything fails
• Fast response over prevention

Continuous Deployment means in practice

• Deploy new software quickly (IMVU time from check-in to production is 20 minutes)
• Tell a good change from a bad change fast
• Revert a bad change quickly (and shut down the line)
• Work in small batches (At IMVU, a ‘large batch is one coder’s code for three days)
• Break large projects down into small batches

Minimum Viable Product is fine. Why? If you are a startup the only people who will talk to you are early adopters. Visionary customers will fill in the gaps on missing features if the product solves a real problem for them.

MEASURE

Rapid Split Testing. A/B testing is key to validating any action through measurement. The three As of measurement: Actionable, Accessible and Auditable. Measure the macro. Only ship things that make customer behaviours better. over time, your product will be more effective, even if it is not prettier.

Eric’s book is coming out in a year or so. The publishing industry doesn’t do lean startups.