Steve Blank is among the most respected authors and teachers in the world on the topic of modern startups. His work at Stanford, Columbia, and NYU is a global model for entrepreneurial education. Beyond that, the U.S. National Science Foundation Innovation Corps uses Steve’s approach as an important foundation for technology transfer between the federal government and the private sector.
Blank’s approach to startups and innovation relies on concepts such as “evidence-based entrepreneurship” and “customer development.” These ideas reflect the belief that startups are “temporary organizations designed to search for repeatable and scalable business models.”
These concepts lead to several implications:
- Startups should base their search for a viable business model on substantive data gathered by talking directly to potential customers
- Iterating on experimental business models is essential to developing a successful startup. Blank relies on the Business Model Canvas from Alex Osterwalder.
- The “minimal viable product” enables a startup to experiment, test assumptions, and then adapt quickly based on market feedback
These (edited) comments from Steve Blank help illuminate his advice to startups and his approach to innovation in large organizations:
On gathering data by talking to many potential customers: When you get out of the building and talk to people, you will likely find that your initial hypotheses are wrong. So, in a lean startup we get out of the building to accumulate real data and evidence very early. I’m talking about good enough decision-making not based on guessing, but on the fact that we are talking to a ton of people.
On failing fast and pivots Startups and new ventures go from failure to failure. That is a huge concept. You can count on two fingers the number of start-ups that end up with their initial plan.
When you make a major change to any part of the business model canvas – whether it is changing channel, customer segment, etc., we call that a pivot. A pivot is a substantive change to any component of the business model canvas based on customer data and real evidence. Not one data point but a bunch of experiments.
Iterative experiments could be a PowerPoint of an image of the product; it could be a wireframe; it could be a clay model of a heart valve or a catheter. It could be sampled data for a therapeutic drug that you think you might get back from clinical trials to show a potential pharma partner.
Unless you understand all those questions, you will have neat technology and a failed business.
On minimal viable products and agile engineering: We call these incremental and iterative prototypes minimum viable products. There is a series of them and they change over time. They could just simply start with a PowerPoint or web page that says, “If this is important, click here;” to a wireframe; to a model; to a prototype; to eventually low-fidelity and a high-fidelity versions of the product.
[By] constantly iterating, there’s no alpha test, beta test, first customer ship – that’s a waterfall model. We use this agile engineering model, but we couple it with the customer development philosophy and the Business Model Canvas. We use agile engineering to test all the hypotheses of the business.
On corporate innovation and startups: Companies are looking at startups to do continuous innovation. For the first time, large corporations realize they need to move at a speed they never had to before, because historically they could own a market segment and defend it for a decade or more. That’s almost impossible now unless you own the regulators and/or politicians. In almost every other market, you have to figure out how to innovate at a very different speed.
On the challenge of corporate innovation: Here’s the big idea: in a corporation, we have spent the last century building execution tools, processes, and procedures that allow us to execute efficiently. It turns out that the tools and strategies, policies, and procedures to innovate are the antithesis of those execution tools. And that sets up a real conundrum, “How you get the two to operate in parallel without screwing up the ability to make money?” That’s a real interesting problem.
On the question, “Does execution efficiency kill corporate innovation?” Kill is an understatement; it strangles it in the crib! And I don’t mean with any malice; but by default. Clayton Christiansen, whom I think is the direct descendent of Schumpeter in terms of strategy and understanding this in corporations, articulated it brilliantly in the Innovators Dilemma, in the 90s. I think we are still working out some of these processes.
It has taken the tactics of lean start-ups to articulate the actual tools that we would use to solve this problem.All the things we do: the HR manual, the comp plan for sales; the branding guidelines — just make the list — everything puts a roadblock into innovation that startups don’t face, and that’s the irony. That’s why startups move so fast.
CxO-Talk brings together prominent executives, authors, and analysts to discuss leadership, transformation, and innovation. Join me and Vala Afshar every Friday for a new episode of CxO-Talk
(Cross-posted @ ZDNet | Beyond IT Failure Blog)