Global retailers like AutoZone are using big data to customize selection at the local retailer level. For example, AutoZone customers in Waco, TX might find a deal on shocks that wouldn’t be available at other AutoZones, while customers in Mulberry, FL might find a special deal on bug deflectors. AutoZone is adjusting its inventory at some of its 5,000 locations based on information gleaned from multiple databases. There are a number of startup database companies now appealing to large retailers’ needs for a faster and more efficient approach to data collection and management.
Retailers are now relying on “Big Data” to capture more customer information. Big data is defined as data sets so large and unwieldy that it is difficult to manage, process, and capture data on traditional database systems. Big data allows analysts to explore relationships beyond basic correlations. For example, Google can better predict a flu outbreak than the CDC using big data gathered from customer search terms. Using big data, Walmart found that during storm preparation, customers bought flashlights and pop-tarts, although no causal relationship exists. Big data also helped analysts realize that credit scores can help predict who needs a reminder to take medicine.
Google, Amazon, and Facebook have been using big data for a while. Chevron uses multiple databases to process seismic images to search for new reserves of oil and gas. Sears and Walmart now use Big Data to aid in their marketing efforts.
The market for big data databases is worth $1.22 billion and is expected to double by 2014. In the past two years, 119 database software companies received $1.17 billion in venture capital funding. Conventional databases are written in SQL, but these databases can’t handle the large quantities of data generated by social media, mobile devices and other technologies. Only recently have research firms been able to capture unstructured data from the Internet.
Many of these new databases can be accessed over the cloud and don’t have the high maintenance or access costs of conventional databases. Other databases only charge a consulting fee. NuoDB, a startup database service, uses a cloud with an annual subscription fee. AutoZone uses NuoDBto quickly analyze data without freezing its system.
However, not all large corporations are quick to adopt this technology. For example, some financial services providers are sticking with conventional relational databases to speed up transactions. Some fear that with the new databases, there are higher risks of mistakes. Oracle, the largest database vendor, is developing new databases that no longer run on SQL. Online real-estate site Trulia uses one of Oracle’s open-source conventional relational database systems. However, as it began to grow, Trulia experimented with other databases to process some of the information on the 100 million U.S. homes in the company’s system.
1. What is Big Data?
2. How is Big Data used by retailers?
SOURCE: Rachael King and Steven Rosenbush, Wall Street Journal, March 13, 2013