Visionary vs Data-driven Development(aka why Lean is not a one-size-fits-all solution)

By Jeremy Schreiber

In class, we study the lean-startup method, and with that we examine a very data-driven approach to building a product and a company.  While this is a very effective way to design for a consumer-facing online platforms, where scaling rapidly (or virally) is essential, I still question this approach as a one-size-fits-all solution to building a startup.

As discussed in class, the visionary vs. data-driven development problem can be easily seen by comparing Apple and Google.  Google is one of the most data-driven companies around.  They design and quickly launch beta products, getting the code into the hands of their users as fast as possible, and learning from their observations.  And yet the solutions they create, while diverse and powerful in their own right, are fragmented, messy, sometimes unnecessarily complicated, and lack a cohesive feel.  Many of their products solve existing problems quite well, but perhaps with the exception of Google Maps and Search’s autocomplete, fail to wow most users.  Apple on the other hand builds all of their products based on the vision of a few elite designers (namely Steve Jobs and Sir Jony Ive). Almost every product they’ve introduced in the last decade has blown users’ minds and far exceeded expectations.  So is there a lesson there for startups?

I believe that to build a good product and a good company, the founder should be a true visionary.  That means seeing things that other people don’t see. Only in that circumstance can a founder truly have the passion to build an amazing product.  This doesn’t imply the founder has any magical powers of foresight, but perhaps rather that they have an innate ability to empathize with people.  Design thinking principals teach us to learn from others by truly understanding them.  Simply asking them questions about their preferences and collecting data is not enough.  You must observe the way they live their lives, walk a mile in their shoes.  Visionaries are those who can understand people better than they know themselves, and fill a void in their lives that they don’t yet realize is missing.  Once this void has been recognized, the founder will find him/herself building a solution to a real problem, rather than creating a solution in search of a problem.  At that point, a Steve Jobs-esque reality distortion field is no longer necessary. Unfortunately, no amount of lean-startup hypothesis testing will lead to the discovery of such a vision. Only the powers of empathy and human observation will lead to such breakthroughs. 

So I sometimes feel that the data-driven, hypothesis testing, lean-startup framework, when used as a fundamental approach to starting a company, is most useful when a founder is not a visionary, but rather a person who is just excited to start a company.  This also holds for those that may have a rough idea, but are trying to find an angle to make it work. But the best ideas, products, and companies, are born out of necessity, passion, and vision, not out of rapid hypothesis testing and user A/B tests.  Think about the core technology innovations that have changed the world:  microprocessors, MSDOS, transistor radios, LCD Monitors, touch screens, iPhones, etc – these were not built on lean principals, but rather on the passion of technology visionaries. 

Perhaps I am a bit biased as a former hardware engineer, but I feel that lean principals apply best to a world of proliferating software, where there is little to differentiate one product from the next. Here the utmost care must be taken to ensure that every dollar of development effort is used towards learning and bringing software to market as fast as possible and in a way that meets the needs of as many users as can be done given the circumstances. In this scenario, lean principals help to deliver better software solutions and minimize aimless development towards ambiguous goals.    

That said, even within the realm of a visionary founder with a visionary idea, lean principals can still prevail as a methodology for developing the product as efficiently as possible, and helping to ensure that the business model around the vision evolves quickly.  So I will argue that while lean-startup and other data-driven methodologies are not a one-size-fits-all solution to developing a startup, they can be a helpful set of tools in the quiver of any successful company.