Eight years ago, business intelligence was red-hot. Data Warehousing, OLAP, ad hoc reporting were the in buzzwords, and for good reason. With all that ERP data available now sitting happily in corporate databases, and with computing power and storage space becoming cheaper and cheaper, the natural next step for IT was to find something to do with it. (Let’s ignore that whole decision support period in the 90’s and pretend that the business computing space isn’t thematically cyclical).
Meanwhile, companies such as SAS and SPSS with roots in research computing and driven by leadership who embraced mathematics worked in a field that could be called predictive analytics. (I’m not clear what they really called it in the 80s and 90s – but if you needed to run a statistics program in graduate school, there’s a good shot you used SAS or SPSS). Long before cheap hardware made it possible to record and report every single business decision and event made by every employee in a company, predictive analytics practitioners concerned themselves with surveys and sampling and statistics – asking questions, setting up business experiments, and trying to project their results into the future. Now, it’s true that SAS was used pretty heavily for basic reporting in a lot of places, but at the end of the day, the things that SAS is good at could never be done with a simple *cough* reporting tool.
It’s all Business Analytics.
Lately, that division isn’t so clear. Tools that were branded as business Intelligence tools are slowly becoming business analytics tools. IBM has purchased SPSS, giving them industry leading solutions in both business intelligence (Cognos) and business analytics (SPSS). SAP BusinessObjects has rolled out a new SAP Predictive Analysis tool, including integration with the R statistics platform. Several major consulting firms, including Deloitte and Accenture, are advertising their Analytics practices. Is there a true difference? Is this a change in product or a change in marketing? It’s both.
Google Trends Report: Business Analytics (blue) vs. Business Intelligence (red)
Google Trends Report: Business Analytics (blue) vs. Business Intelligence (red)
Analytics is the art of data-evidenced business decision making; analytics tools are used to collect, explore, and analyze business data. I like to think of it as one space with three primary focuses…like this.
| Business | Analytics | |||||||
| Data Collection | Data Exploration | Advanced Analysis |
||||||
| Data Integration | Reporting | Data Mining | ||||||
| Data Cleaning | Dashboarding | Predictive Modeling | ||||||
| Data Warehousing | Visualization | Business Experiments and Optimization | ||||||
Now, what do I get for my money?
Business Intelligence: Measuring What Your Business Is Doing
Business intelligence techniques, listed above as data collection and data exploration, help business leaders know *what* is happening in their enterprise by directly measuring the data captured in their business systems. Need a list of outstanding invoices? Need a data mart to combine information from your accounting system and your time reporting tool? Need to clean up that vendor file? Those are Business Intelligence techniques and until very recently have been the bread and butter offerings of companies like SAP BusinessObjects, Informatica, and IBM (Cognos). In general, business intelligence is pretty well understood in the marketplace today so I’ll go a little thin on the details and skip to the new stuff.
Advanced Analysis Techniques: Understanding Why and What to Do About It
Newer business analytics products and techniques (or in some circles, data analytics or predictive analytics…or ten years ago, data mining) help business leaders understand *why* their company perform the way that it does and to find ways to change that performance. The good news to consumers is that, for the most part, they use pretty standard mathematical techniques for a common set of core capabilities and tend to be interchangeable. In addition to generalized analysis toolsets, several firms – including the large toolset vendors – also offer industry-specific modeling and analysis tools that address common business problems. These are a little more varied…
Traditional Business Analytics Capabilities
- Data mining, categorization, and classification
- Forecasting and predictive modeling
- Business experiment design and execution
- Process optimization and operations research
Applied Business Analytics Examples
- Market segmentation and product development
- Credit risk and fraud detection
- Retail and market basket analysis
- Supply chain optimization
- Financial analytics
What’s next?
At this point, business intelligence and analytics capabilities are pretty easy to obtain, if not always cheap. (I’ll write later about open source analytics, and mixing and matching components to build a right-priced solution…) So what’s next? That one’s clear, and the topic of a future post: Big Data.