A sustainable competitive advantage is the ability for a business to do something that its (potential) rivals cannot. It means being able to:
- Keep the business’s current customers; and,
- (Possibly) grow by attracting new customers;
- On terms which are more profitable than the business’s competitors;
- For a reasonably long-time (it is sustainable not temporary).
I think the best competitive advantages in achieving this are, in descending order:
- Economies of Scale combined with Customer Captivity;
- A Network Effect;
- A Government Licence;
- Patent Protection; and,
- Customer Captivity. 
How a business gets a competitive advantage
A business usually gets a competitive advantage from deliberately trying to gain a competitive advantage, or simply by good luck.
Where do you find businesses with sustainable competitive advantages?
I agree with Greenwald and Kahn’s conclusion that a business with a competitive advantage is much more likely to be ‘local’ and operate:
a) In a limited geographical area (Wal-Mart in the US south); or,
b) In a limited product space (Microsoft with its operating systems).
This usually applies irrespective of the type of competitive advantage. For example, a business with an economies of scale competitive advantage is more likely to be ‘local’ for two reasons. Firstly, operating in a small market makes it is more difficult for a new entrant to capture enough market share off the incumbent to reach economies of scale. Secondly, high fixed costs, the linchpin of economies of scale, are fixed only within the region or product space in question.
A business with a customer captivity advantage is also more likely to be ‘local’ because it is often selling a customised product or service to its customers. Providing this customised product or service usually needs the business to specialise in a limited geographical area or product space.
The restricted nature of a government licence or patent also means that businesses with these competitive advantages are often in a limited geographical area or product space.
A competitive advantage is limited to the product
If a business has a competitive advantage with a particular product that does not mean it has a competitive advantage with its other products. For example, Pepsi drinkers have no particular attachment to KFC.
I will discuss each of these different types of competitive advantage, and how strong I think they are, soon.
 Other than customers that naturally die or age out of the market.
 Customer captivity refers to how a business has ‘captured’ a customer. This can be because the customer will have high costs in ‘switching’ to an alternative product or an alternative service provider, or high costs in ‘searching’ for this alternative, or simply because the customer has developed a strong ‘habit’. I will discuss this form of competitive advantage in much greater detail in a later post.
 I think customer captivity is not as powerful as the other competitive advantages because it only allows a business to profitably grow from its current customer base. It does not allow the business to grow by attracting new customers on the same profitable terms.
 A business may have a competitive advantage simply because of good luck. As Denrell points out: “(it) is often argued that a strong performance record by a business cannot be because of good luck but indicates superior practices. The underlying idea is that sustained high performance is less likely to be the result of chance events. In other words, high performance during a number of years will filter out random noise and better reveal average tendencies, perhaps determined by capabilities. The reasoning assumes that the performance of an organisation is the sum of independent and random variables. If the random variables are not independent, high performance during the whole period may not be very impressive. An analogy makes the point. It a runner wins 10 independent races, she is probably a good runner. Suppose the races are dependent, however. That is, if a runner won by one minute, she starts one minute before the second runner in the next race. Clearly winning 10 such races is less impressive because if you won the first, you should have a higher chance of winning a second… Many industries consist of such dependent races.” See, P 295, ‘Should we be impressed with high performance, J. Denrell, Journal of Management Inquiry, September 2005.
 Ibid at P3.