Analytics is the new buzzword. There’s big data analytics and advanced analytics. Predictive analytics is moving swiftly along Gartner’s 2011 Hype Cycle for Emerging Technologies towards mainstream adoption. Social analytics is also moving upward on the curve. All in all, business analytics (BA) is in. But does that mean business intelligence (BI) is out?
Not according to Sean Paine, chief operations officer of information solutions specialist, EnterpriseWorx.
“It’s fashionable to talk analytics right now,” he says. “But you need data warehousing and business intelligence as the information backbone for analytical applications. All three elements must work in unison to gather, store, mine and analyse data so as to guide business decisions.
“BI encompasses storage of data and the technical engines for making sense out of billions of rows of data to produce one or two meaningful dashboards. As well as analytics, operational efficiency and business strategy rely on information from BI.”
As BI and BA evolve rapidly to guide decisions in areas from marketing, research and development and inventory management to customer care, so business analytics will become more prevalent.
BI and BA are separate but connected tools, according to Paine. “BI is about being able to drill through data to find information – usually through querying, reporting, online analytical processing (OLAP) or other straightforward analysis tools,” he says. “BA and advanced analytics tap into statistical and quantitative data for explanatory and predictive modelling.”
According to The Data Warehouse Institute researcher, Philip Russom, “Advanced analytics is very much about open-ended exploration and discovery in large volumes of fairly raw source data. OLAP is about a more controlled discovery of combinations of carefully prepared dimensional datasets.”
“BI provides the context for BA,” says Paine. “It can explain what happened, where, as described by existing measures and dimensions. However it takes BA to address more complex questions such as what increases sales? What predicts customer loyalty?
“In short, OLAP is a way of making sense from a lot of data. BA is about looking for relationships once you’ve made sense of that data. BA uses statistical analysis to identify causal and complex relationships. These inferences and feedback can then provide input into the business decision making process.
“Unstructured data is more amenable to BA than to more structured OLAP tools. This is because BA can uncover patterns and relationships in the big data sets generated by social media. You could use analytics tools to analyse sentence structure and frequency of certain words in customer complaints, for example.
“Then, by feeding that information into the OLAP system, it would be possible to work out the education level of your consumer base, so that the marketing team could work how they should be pitching their advertising.”
New data-management approaches are challenging the traditional process of delivering business intelligence solutions via OLAP cubes built on top of data warehouses because of latency in reporting and volume restrictions.
One new approach for avoiding such bottlenecks, such as Kognitio WX2, uses a combination of in-memory power and massive parallel processing (MPP) to enable complex analytics on terabytes of data in seconds.
“Connecting the BI or BA front-end to the existing data warehouse remains a challenge, best left to experts with the relevant expertise in data integration,” says Paine.
“From the user point of view, there is a move towards providing deeper insight into business information. Putting powerful tools in the hands of managers on the front line of business means that they can use these solutions to make long-term decisions.
“BI and BA are no longer tools for the exclusive use of IT specialists and analysts, but can be used to guide business strategy. BA, in particular, allows the organisation to move from describing the past to predicting the future.”