Big Data analytics go far beyond just analyzing very large datasets. They center on use cases that require the integration and analysis of different data types and sources. These range from marketing use cases such as advanced web analytics, online product usage (cohort) analysis and social media sentiment analytics, security and risk use cases such as fraud detection, identifying rogue trader activity and asset risk analytics, and scientific or research use cases in medical, pharma and related industries.
With traditional business intelligence tools, users are only able to analyze structured data, which limits the amount and kinds of analysis they can perform on their data. With big data analytics, it is now possible to quickly bring together and work with all types of data from any number of data sources, whether it’s structured transaction data or semi-structured or unstructured data such as weblogs, social media data or emails. Analysis of all data helps users understand and analyze both customer transactions and interactions and lets them answer questions and find insights that simply are impossible from traditional BI and structured data alone. For example, companies want to understand the entire customer engagement cycle from a web ad all the way through to a purchase, either online on in a store, so that they can see just which web ads actually result in the highest purchase percentage, rather than just the highest click rates. This requires the integration and correlation of weblogs, clickstream analytics and transaction data in order to get the complete picture.
