Definition: Big Data
Many experts refer to data as the fourth factor of production, following capital, labor and resources. This change is logical as data gains more and more importance. Current estimates suggest that the volume of data generated worldwide will even double every two years. In order to use data as a competitive advantage, companies first need to link it and then analyze it. These methodologies and technologies fall in the category of big data.
Risks of big data
The largest risks involved with big data are typically inadequate technical and business knowledge as well as an acute lack of qualified personnel and, therefore, understanding of the technical details. This often leads to misunderstandings, bias, rejection or the insufficient use of big data – and, therefore, companies drowning in the flood of information. The underlying reasons often have to do with data volume, velocity and variety. Experts have identified these three ‘V’s’ as the main challenges and problems associated with big data projects.
The prerequisites for successful big data projects are suitable appliances, analytic databases or a cloud architecture, which make it possible to process such vast amounts of data. Examples include the data platform SAP HANA or the Parallel Data Warehouse (PDW) of Microsoft. What matters is the efficiency – and not the scope – of the tools used for the solution. The best system, in other words, is not necessarily the most expensive one.
Whereas the U.S. market tends to focus on the opportunities of big data, there is a growing concern in Europe – in particular, Germany – that big data breaches data security laws and bears the risk of uncontrolled surveillance. Using big data for personalized advertising is also an ethical question. Interlinked customer data can lead to an invasion of privacy and, therefore, result in data privacy risks. Careful, responsible usage of data is just as essential as communications that clearly convey that customer data is anonymized and will not be misused for advertising purposes.
Big data use cases
Just imagine: face-recognition software in a beverage vending machine that automatically recommends a suitable drink for the person standing in front of it. This big data use case is already reality (http://derstandard.at/1381370985196/Gescannte-Kunden-fuer-gezieltere-Werbung). That the drink prices at the vending machines vary based on the facial expression that shows how thirsty the individual is, however, is still just a rumor.
Real-world big data use cases include staff scheduling at a drug store chain, price optimization, credit card fraud protection and R&D.