by John Bacon
I’m a lucky guy! I have worked for very large high-tech manufacturing companies, led software company public offerings both in the U.S and in Europe, and co-founded my own company. Plus, I am faculty for the National Science Foundation’s I-Corp program.
Some of you may know I-Corps as the result of an audacious initiative between your federal government and Steve Blank, serial-entrepreneur turned academic, and the thinker who launched the Lean Startup movement. We have now admitted to the I-Corps program almost 600 teams from the best research labs in the nation to teach them the tools of Lean Startups. Top scientists who survive the boot camp experience come away with the skills necessary to pinpoint the value proposition provided by their technological breakthroughs, exactly who will buy the associated product, for what reasons, and in what business model.
I have been involved with the program since the beginning (4 years) and have worked directly with Steve Blank, NSF, and university leaders to bring it to life. I teach several cohorts every year, and I interview new applicants each week. The results have been amazing, and you – as taxpayers – are beginning to get much improved returns on your federal research tax dollars.
So it makes sense that large organizations are looking at the Lean Startup model (LSM) to see how it can be adopted for their use, a place where I also have experience. But the transition of Lean Startup practices into large organizations isn’t simple!
As I work with large companies, I observe emerging patterns which need to be studied, since – as with any new idea – mistakes and misapprehensions abound. Here are some of the more common ones:
Good stuff . . . we’ll integrate it into our stage gate process! Peter Koen made it crystal clear at the IRI Member Summit in Chicago that you are doing the wrong thing if you attempt to stretch your stage-gate processes into the Front End where transformational innovation is born. The process to use for that is LSM. Don’t conflate the two. Stage-gate is for Sustaining Innovation initiatives, NOT for Front End projects.
Well, we have great marketing and business development people who know our R&D needs; we’ll use them to form a team and get off to a fast start! Nope. Corporate marketing people – good as they may be – are genetically incapable of doing the Customer Discovery required to make LSM work. Why? They will always use proven Voice of Customer (VoC) techniques instead. VoC filters all the potential data to be gathered through the corporate lens of products, markets, business models, and how-we-do-things-today. This unavoidable, current-state focus will corrupt the data. Customer Discovery can only be done by a carefully selected and trained SWAT team who can free themselves from all the corporate-ness and go out to talk to human beings about NEW IDEAS the humans (not the marketing people) care about.
But, we have to conform to our corporate strategy road map, so we need to validate the product features which are planned! . . . which means that you’re not in the Front End; go do your VoC and stage-gate magic. At the core of LSM is the scientific method. Richard Feynman did pretty well by guessing (creating hypotheses) and then using experiments to confirm or disprove them. The irony is that your R&D colleagues know this approach well, but they don’t know it applies to learning new insights about business and customers, not just science. LSM – properly implemented – uses the same approach to quickly evaluate whether the intended customer cares enough about your new ideas to justify the investment to form what will become a sustainable business. Using properly constructed hypotheses, it allows the rapid iteration required to gather real-world feedback and learning (insight) not previously understood by your organization.
We know our business; we’re going to go straight to a Minimum Viable Product (see? we have read all about LSM!) and accelerate the process! Customer Discovery is scary, messy stuff, so teams are tempted to believe they can “sanitize” the process by using a cool prototype to make things more comfortable; after all, a prototype lets your colleagues stay in their comfort zone and talk speeds and feeds! Unfortunately, this never works. Customer Discovery data needs to drive development of the Minimum Viable Product at least after the first crude version. Why? There are three reasons.
First, a polished and complete prototype project is what YOU want; it excludes the valuable involvement of the customer and their buy-in to creation of a future product that they feel ownership in. Second, a fancy MVP will bring in your lawyers and their 5 page NDAs which will kill the conversation you want to have. LSM is about learning unmet needs of specific customer archetypes, and that conversation doesn’t require you to risk crossing the publication bar or giving away trade secrets. Finally, speed is critical in the Front End. Your MVP should be made of clarinet reeds, Scotch tape, some Lego bricks and a toilet paper tube. Literally. What you’re trying to do is get the customer to say “No, the superlator nozzle tube should go on the other side because. . .” That’s where insight begins. The fancy chrome plated version that took eight weeks to build in your 3D printer lab will only impress your hard-won interviewee into silence.
We’re product guys; we can’t fool with our corporate business model. Adherence to the existing business model is one of the biggest killers of real innovation. One of the best parts of the LSM method is the construction of the Business Model Canvas from Alex Osterwalder. If you truly believe you must stop short of the full LSM process by trying to shoehorn your cool new ideas into the existing model, you may need to recruit a senior executive from another area who can support the changes (something you should have done to begin with on this and every FEI project!).
We’re ready to go with Customer Discovery; we got the list of names from corporate marketing! That’s the wrong list. I’ve seen lists that included the names of companies (not humans), distributors (not your customers) and purchasing agents (have you ever seen a PA buy something new and different?). Notice that it’s called Customer Discovery, not improvement or refinement.
The LSM model begins as a series of hypotheses designed to test new value propositions against very specific customer archetypes. A great idea tested against the wrong customer (or an entity that is not a customer at all) won’t get you very far.
Gadzooks! This isn’t easy! Correct. But, just as with any other hard, valuable skill you’ve acquired, it gets easier with outside coaching and practice. Put in the time to learn how to do it right, and the Lean Startup Method will reward the work!
It will be worth it!