The sales and service network planner and the CFO at Tesla have a much harder job. Tesla owns all of its stores, it has to pay for land and building, buy tools, lifts, and paint booths, it must meet the payroll of each of its stores, no matter how many cars are sold. From a network planning standpoint, hypergrowth is an invitation to bankruptcy if you have to finance the stores all by yourself. Let’s get the back of an envelope, and run a few numbers.

According to Tesla’s website, the company has a total of 226 fully owned Tesla stores in North America and a few foreign countries. (Please do not blame me if I haven’t counted stores that are not on the website. Tesla has been asked to verify the number, but did not respond. See below.) Using Tesla’s 2016 target of around 80,000 cars, the average Tesla store turns 354 cars a year. That already is a very bad key ratio. The average U.S. franchised new car dealer moves close to 1,000 units per year. Most large automakers would try to fire a 300 unit dealer, or if that fails, drive him into bankruptcy, just because they don’t want to bother with such small fry.
Now let’s assume that Tesla is serious to sell 500,000 cars a year by 2018. How many stores would that need? If Tesla continues on the level of current inefficiency, it would need to own and operate more than 1,400 Tesla stores worldwide by 2018, and double that two years later.

source - http://www.forbes.com