I'm not sure of the statistics, but based on the people I know, it seems that most businesses are started by people with previous experience in that industry. You work for a law firm, then start your own. You sell cars for someone else, then take out a loan and open your own dealership. You write code for Google, then leave to build a new web application. That's the way it usually seems to happen. But is that the way it has to be?
An article asking How Relevant is Drucker, made an interesting observation about Indian business culture that highlights a major difference with the West.
At a recent SME (small and medium entrepreneurs) conference, three leading Indian businessmen echoed a similar thought. Their choice of industry was not something that they knew anything about, but consciously one that they knew nothing about. They reflected that this was because it ensured they worked hard to learn and listen to others and perhaps their "naiveté" ensured that there was enough innocence left to provide for innovation. This goes against classic western business management theory of driving business by core competence.
What does that say about core competence? Should a new startup be based around some sort of competitive advantage that stems from previous experience? I think most investors would expect that. To trumpet inexperience as your advantage is one way to make sure you get turned down for funding.
Yet, at the same time, there is something to it. Existing companies are often in such a routine and do so many things out of habit, that they miss opportunities that should seem obvious.
Back in college, one of my professors had worked at Bell Labs, and had received a patent on a circuit he designed his first year on the job. He said that he did it because he "hadn't been around long enough to know that it was impossible." His co-workers had let design assumptions slide by without questions, even as tools, industries, and business models had changed.
So where does that leave the entrepreneur?
Statistics usually trump anecdotes, so I think in most cases, it is best to stick with industries you know. It's easier to see opportunities, you have a stronger network of contacts, and you have a knowledge base to build on that gives you an advantage over competitors. But there are exceptions. There seems to be a rare type of entrepreneur whose skill set is nothing but pure entrepreneurship. Such people possess functional knowledge about evaluating markets, pulling together resources, making the most of limited capital, and managing the process of new company creation. They probably have functional knowledge related to certain industries too, from time spent in a more "regular job," but the key difference is in the industry and business meta-knowledge they acquired that most people never pay the slightest attention.
As with most things entrepreneurial, the best advice is to know yourself well and understand your skills and competencies. Some industries are easy to learn. Some business models leave lots of room for a learning curve. Some industries will kill you for the slightest mistake. Being blindfolded isn't a bad thing if you're listening to music, but if you are driving a car, it's deadly. Startups are the same way.