The most common sources of business risk seem to stem from excessive leverage. When a company owes more money than it is worth, it is unlikely to survive. If you're in a business that depends upon making loans, and you are also leveraged, the risk to you if your customers default is magnified many times. In the United States, before the financial crisis of 2008, banks were allowed to borrow $33 for every $1 that they held in equity. So imagine a bank that could borrow money at 2% and lend it out at 3%. If they had $10 million in equity they could borrow $330 million in debt, lend out that money at 3% and make a gross profit before expenses of 3.3 million per year. Or a 33% return on their 10 million in equity. However, only a very small portion of their loans needed to become unrecoverable or in default for the bank to become insolvent. 10 million out of 333 million is only 3%. Another type of risk is reputational, damage to your brand and ability to sell in the future. One area where any restaurant chain, manufacturer, grocery store, or retailer is extremely vulnerable to bad publicity and damage to reputation is when they have sold a food product that is tainted or contaminated and makes customers sick. One successful example of risk management using big data in the product recall space is Costco's rapid response to product recalls. The metric for Costco is how long it takes them to contact every potentially at-risk customer, when what's called a level one product recall occurs. Because Costco requires all of its customers to purchase an annual membership, and their membership database contains address information and often email and mobile phone numbers as well, it's relatively easy for Costco to track down customers. This is easy because Costco stores a record of every purchase made by every customer at the individual product ID, or SKU level. So that data is indexed and when a product is recalled by the manufacturer, as happened twice in the spring of 2014, with peanut butter and again with peaches. Within 24 hours, Costco was able to identify all of the more than 50,000 customers, who had bought the product and used the fastest available means to contact them, in some cases email and phone calls, but in any case a letter to a mailing address, and they were able to prevent many, many people from getting sick. An even more interesting example of using Costco's database to track an epidemic to its source occurred in January, February of 2010, when 272 people in 44 different states became ill with a single genetically identical strain of salmonella. United States Centers for Disease Control, or CDC, sought to track down the source of contamination, but were initially unsuccessful. Until someone had the very bright idea of analyzing Costco purchase records. For those in the group of 272 who might have shopped at Costco. Since Costco membership reaches about one-quarter of US families, it would be reasonable to assume that this was about 70 people. It was then possible to identify what food items those 70 had purchased in common. The only common product was pepper salami, manufactured by a small sausage company, Danielle International, based in Pascoag, Rhode Island, and the contaminant turned out to be one single 25-pound box of crushed red pepper provided by wholesaler, wholesome spice, based in Brooklyn, New York. Not so wholesome after all.