Editors note: This is the first part in a multi-part series on analytics.
One of the hottest technology topics today is analytics. The problem with analytics is that few people agree what it is. This often happens with commonly used terms because everyone attaches a slightly meaning to them based on their needs and perspectives.
I prefer to assign two definitions to analytics to reflect the primary dimensions of the term: its industry context and its technology context. For simplicity's sake, Analytics with a capital "A" is an umbrella term representing our industry, while analytics with a small "a" refers to technology used to analyze data.
Analytics With a Capital "A"
Analytics as an umbrella term refers to the processes, technologies, and techniques that turn data into information and knowledge that drive business decisions. The cool thing about such industry definitions is that you can reuse them every five years or so. For example, I used this same definition to describe "Data Warehousing" in 1995, "Business Intelligence" in 2000, and "Performance Management" in 2005. Our industry perpetually recreates itself under a new moniker with a slightly different emphasis to expand its visibility and reenergize its base. (See "What's in a Word? The Evolution of BI Semantics.")
Today, many people use the term Analytics as a proxy for everything we do in this space, from data warehousing and data integration to reporting and advanced analytics. The most prominent person who defines Analytics this way is Tom Davenport, whose terrific Harvard Business Review articles and books on the subject have prompted many executives to pursue Analytics as a sustainable source of competitive advantage. Davenport is savvy enough to know that if he had called his book "Competing on Business Intelligence" instead of "Competing on Analytics", he would not be the industry rock star that he is today. (I personally still prefer the term "Business Intelligence" because it perfectly describes what we do: use information to make the business run more intelligently.)
Analytics with a Small "a"
This leaves the term analytics with a small "a" to describe various technologies that business people use to analyze data. This is a broad category of tools that spans everything from Excel, OLAP, and visual analysis tools on one hand, to statistical modeling and optimization tools on the other.
One way to segment analytical tools is to show how they've evolved over time, along with reporting tools. Figure 1 shows that we've had four waves of business intelligence tools since the 1980s. Specifically, there have been two waves of reporting followed by two waves of analytics. The first wave of analytics took place in the 1990s when business analysts began using ad hoc query/reporting and OLAP tools to explore historical data in a data mart or data warehouse. The second wave of analytics, which just began, involves modeling historical data to optimize the present and predict the future. Most people who talk about analytical tools today refer to this latter type.
Interestingly, each wave of analytics follows a wave of reporting. This makes sense if you consider that reporting tools are primarily designed for casual users, who comprise 80% of all BI users, and analytical tools are primarily designed for power users, who constitute the remaining 20%. These are two separate, but inter-related markets, which BI vendors need to address.
Deductive and Inductive Analytics
The first wave of analytics--which addresses the question "Why did it happen?"--is deductive in nature, while the second wave of analytics--which addresses the question "What will happen?"--is primarily inductive.
With deductive analytics, business users use tools like Excel, OLAP, and visual analysis tools to explore a hypothesis. They take an educated guess about what might be at the root cause of some anomaly or performance alert and then use analytical tools to explore the data and either verify or negate the hypothesis. If the hypothesis proves false, they come up with a new idea and start looking in that direction.
Inductive analytics is the opposite. Business users don't start with a hypothesis, they start with a business outcome or goal (e.g., "find the top 10% of our customers and prospects who are most likely to respond to this offer") and then gather historical data that will help them discern the answer. They then use analytics to create statistical or machine learning models of the data to answer their question. In other words, they don't start with a hypothesis, they start with the data and let the analytical tools discover the patterns and anomalies for them.
Interestingly, our industry's former umbrella terms now refer to categories of tools: data warehousing refers to analytical databases and ETL tools; business intelligence refers to query and reporting tools; and performance management refers to dashboard, scorecard, and planning tools. In time, analytics will be replaced as an umbrella term by some other industry buzzword, and the term will simply refer to deductive and inductive tools, or perhaps just one or the other.
The Value of Analytics
Now that you know what advanced analytics is, the next question is, why should you
Your chief financial officer will be glad to know that analytic applications have a higher return on investment than all other BI applications. A landmark study by IDC in 2003 showed that the median ROI for projects that incorporated predictive technologies was 145% compared to 89% for all other projects. This uplift is gained largely by optimizing business processes, making them more efficient and profitable, according to IDC.
But what kinds of questions does advanced analytics address? There are four major categories:
- Analyze the past. Although we mainly use deductive tools to examine past trends, advanced analytical tools model the past. Some seemingly easy questions can be maddingly difficult to answer because they involve the interaction of so many variables. These include, "Why did sales drop last quarter?"
- Optimize the present. Once we model past activity and understand relationships among key variables, we can harness that information to optimize current processes. For instance, an market basket model can help retailers design store layouts to maximize profits.
- Predict the future. By applying the model (i.e., mathematical equation) to each new record, we can guess with a reasonable degree of accuracy whether a customer may respond positively to a promotion or a transaction is fraudulent.
- Test assumptions. Advanced analytics can also be used to test assumptions about what drives the business. For example, prior to spending millions on a marketing campaign, an online retailer might test an assumption that customers located within one square mile of a big box competitor are more likely to churn than others.
Although advanced analytics can be applied to almost any business function, marketing seems to attract the lionshare of analytical work. Research I conducted in 2007 at The
Data Warehousing Institute shows that five of the top seven applications for advanced analytics hail from the marketing department. These include cross sell/upsell (47%), campaign management (46%), customer acquisition (41%), attrition/churn/retention (40%), and promotions (31%). (See figure 2.)
Figure 2. Most Common Applications for Advanced Analytics
From Wayne Eckerson, "Predictive Analytics: Extending the Value of Your Data Warehousing Investment," TDWI, 2007. Based on 166 respondents that had implemented predictive analytics
In addition, each industry has a handful of applications that are traditional candidates for advanced analytics. (See table 1.)
Analytics is a hot technology these days. But like any hot technology, there are multiple definitions of what it means. Analytics with a capital "A" refers the entire domain of using information to make smarter decisions, while analytics with a small "a" refers to tools and techniques to do analysis. On the technology front, there are two major categories of tools: deductive and inductive. The latter is getting a lot of attention since it's required to optimize processes and predict future behaviors and activities.
Advanced analytics (which is more inductive in nature) offers significantly more value than other types of BI applications because it helps optimize business processes and answer questions that enable the business to analyze the past, optimize the present, predict the future, and test core assumptions. Today, marketing is the biggest user of advanced analytics technologies although its uses spread wide and far.
Posted November 5, 2011 2:41 PM
Permalink | 2 Comments |