Checkers is solved and Chess is nearly solved, but can small business problems be solvable too?
Big data and machine learning go hand in hand. Big data requires machine learning because no human or economical workforce can look at all the data in a database with millions of records. It would take a lifetime to get through them all—a bureau the size of a government tax agency. Luckily, machine learning, via stats and AI, provide a means to make sense of mountains of data without the human cost. This is like creating an inventory of “moves” like in Chess or Go.
One of the problems in the space of big data/machine learning is that we are not really ready for these applications. Data science is starting to make its mark in some industries, like mining and oil and gas, but making the transition to applications that people can use to solve consumer problems or business problems, is a big leap.
We think that for conventional business types (i.e., excluding disruptive innovations), valuing a small business opportunity is solvable with data analysis and mining techniques. Our mission is vibrant cities, and our means is small business creation and survival. The algorithm sifts through millions of records about cities and businesses, and computes an number we call potential yearly earnings. For example, a pizza parlor may get a PYE of $20,000 to $30,000 in some Michigan towns. A low number means that Michigan towns may already be saturated with pizza. It also likely means that compared with other types of businesses, pizza parlors have a low expected value. Perhaps the entrepreneur can look to nearby cities and related business types to look out for better prospects. Otherwise, the entrepreneurs has to pursue a highly innovative strategy to succeed, or accept low profits. By contrast, if a dog grooming shop could make $120,000 in profit in some Michigan towns, that means that compared with other cities, those towns seem to have a need for more groomers.
The estimates will only get better as more data becomes available, and as computing speeds up.