75% of startups fail. What do the other 25% do differently? One approach aimed to make starting a company less risky that has become highly popular in recent years is lean startup. In 2004 Eric Ries coined the term drawing parallels to the lean manufacturing process at Toyota. Lean startup combines agile methods in software development with customer development.
While startups for many decades have started with a static business plan created in isolation without customer insights and before a product had been developed, lean startups acknowledge that all that they have in the beginning are assumptions. They allow themselves a phase of experimentation in search for a viable business model in which they identify pains and needs through listening to customers rather than merely presenting their own ideas. In what has come to be known as customer development lean startups first translate their ideas into business model hypotheses. They then test their assumptions one by one ideally starting with the riskiest presumption. From there they pivot their initial idea by changing one or more of the assumptions and retesting. Only when a business model proved adequate customer acceptance and market adoption they focus on development and make the transition to a company with functional departments.
Product development in lean startups is merely a means to be able to test ideas. What is build is derived from what they want to learn and how that will be measured. They start with a minimum viable product that has no more than the critical features. In contrast to the established waterfall model with a long, linear process and late beta tests, lean startups break down product development into small batches. Often they work side by side on one feature in a cross-functional team and release immediately. Agile development in short repeated cycles gives many opportunities to collect continuous feedback and immediately incorporate learnings in further development without wasting time, money and effort on elements that do not work. Compared to covert product development this carries the risk to reveal confidential details to competitors but lean startups are willing to take this risk for the sake of learning.
Startups with a lean mindset realize that traditional accounting is little meaningful for their purpose. While still in search for a viable business model experimenting with different factors their future is simply too unpredictable for long-term forecasts and milestones. Instead they focus on a relatively small set of actionable metrics and good-enough rather than complete data that directly guides their decision making with the minimum viable product as a baseline. Every action they take from there is an experiment in how to improve one driver of their growth rate. Therefore it is crucial for lean startups to gain a deep understanding of their individual drivers of growth, which in most cases either is a sticky, a viral or a paid engine of growth. In analyzing this data they heavily rely on methods such as cohort analysis and split tests. People tend to be optimistic but lean startups are aware that they need to face the brutal reality and pivot their ideas when their drivers of growth do not go up substantially.
When expectations are not met, employees are not fired but further experiments are derived. Dedicated to building an adaptable organization, experience of employees becomes less important in lean startups. Since the direction of the company is still widely open the ability of team members to collect relevant insights and to act flexibly is most decisive.
Initially developed for fast-growing tech ventures small businesses and large companies can also learn from the lean startup approach. For decades companies have focused on efficiency and cost reduction improving existing business models. Applied within a sandbox intrapreneurs in companies can likewise form teams in search for new business models that supplement a company’s product portfolio to respond to the growing need for rapid adoption to an ever-changing reality. Steve Blank calls for an entrepreneurial innovation-based economy as he argues that “employment growth in the 21st century will have to come from new ventures, so we all have a vested interest in fostering an environment that helps them succeed, grow, and hire more workers.”
Applying a lean startup mindset does not make individual startups more successful. Lean startups fail as often but they fail early when the cost of failure is still low. They do not waste time and resources realizing hypotheses that do not prove right but try to find a viable business model well before running out of resources. Learning from their failures and saving money for what has proven acceptance they reduce their risk of ultimate failure and might in the end build a successful business on something they would have never imagined in the beginning.
This blog post originates in a report for my part-time MBA in Business Innovation and Design at Laurea University based on a HBR article from Steve Blank.