Instead of relying on a few hotshots, take a collaborative approach to developing analytic assets and sharing valuable data across the enterprise.
Too many people are confused by the multiple analytics systems and models that have accumulated within their organizations over the years. It's a problem that hides the incredible value that analytics can provide.
The cure for this confusion isn't as painful as you would think. Businesses must learn to approach their analytic assets more like parts of a portfolio of investments. The goal of the portfolio approach is simple: Focus on the data necessary to answer key questions about your business and analyze only the data you need. With this approach, you bring in the asset "owners" from across the organization who can help you best leverage the required information.
It doesn't matter how many terabytes of data you collect. If those seemingly endless zeros and ones in a vast data warehouse aren't helping you answer the questions that move your company forward, what good are they doing?
Understand the various analytic assets they possess, who owns them within the organization, and the insights each can deliver.
Gain quick, accurate answers to critical business questions.
Determine where best to make analytics investments.
Establish greater discipline in the use of analytics.
What are the pitfalls of not taking a portfolio approach? Let's use the banking industry as an example. During the recent banking crisis, one large institution had an analyst in its credit card business who identified a surprising spending trend. Consumers were starting to spend much less at upscale stores and much more at discount stores. The credit card business within the bank used this insight to adjust its forecasts accordingly and take actions to spur additional spending. However, the insight was never shared with the bank's mortgage business.
Even if the insight was shared, there's no guarantee that the mortgage professionals would have acted upon the insight, since it didn't come from a mortgage analyst perceived to know more about mortgages than the credit card analysts do. In addition, the separate databases and analytic algorithms maintained by each unit may have produced contradicting insights.
With a portfolio approach to analytics both the credit card business and the mortgage business would have access to the same trusted, cleansed, and well-managed data from across the company. They would collaborate on building a series of smaller analytical models focused on their own areas of the business, but these models would roll up into larger analytics models to address questions and decisions facing the entire organization. The company would have a knowledge repository from which all members of the organization could get access to self-service visualization tools that would allow it to see and understand the insights from the various models. The repository would also highlight which experts from across the organization could help other employees with various business challenges.
The portfolio approach hinges on the ability to collaborate, which runs counter to the "hotshot" approach to analytics, in which companies rely on a single data superstar or a limited team of data specialists. Superstars and specialists tend to carve out single-department fiefdoms rather than collaborate across the entire enterprise. These groups may bring value, but the organization will likely have difficulty getting answers to more complex questions that cut across more than one department or that make use of previously untapped data sources.
A hotshot analyst at a telecom company built a churn model with marketing data to predict which accurately customers were likely to drop their accounts within 15 days. The hotshot was proud of the accuracy of the model. Unfortunately, 15 days did not give the company much time to prevent customers from leaving. Months later, a new analyst joined the company and asked, "What if we accept lower accuracy in order to make earlier predictions?"
Armed with this new perspective, the analyst collaborated with operations to obtain data on dropped calls, including details on when calls were on 3G vs. 4G. Using this new data, the analyst built a model that identified (with only slightly less accuracy than the first model) customers who were likely to leave within 60 days. The company had more time to take action, and customer retention soon improved.
A collaborative, portfolio approach gets analysts from across the organization working together, using cross-functional insights to answer discrete questions. For many companies it's a new way of working, demanding collaboration among informational technology (the traditional keepers of data), management (the traditional drivers of business), and analysts (the traditional builders of models). It's an approach that can be applied to a single division or to an entire global organization.
No matter what type of organizational structure you create, it's important to give equal weight to three distinct skill sets. First you need business skills to understand the decisions that analytics should inform. These people need to have skin in the game when it comes to the metrics used to judge the effectiveness of decisions. Second, you'll need analytics and math skills to develop an approach for answering the business questions. Third, you'll need IT skills to provide accurate data for building models and in order to deploy the necessary operational systems to execute the models and decisions.
You can address the most pressing business questions more effectively by breaking down analytics silos and thinking more creatively about how your organization can collaborate across groups. The alternative is a series of separate, meandering trails that more often than not lead to dead ends.
Andy Rusnak is the Americas Enterprise Intelligence Leader at EY (Ernst & Young), where he oversees the firm's offerings in analytics, business intelligence, information strategy, and enterprise performance management. He was previously vice president of Worldwide Technical Operations at Hyperion.
There's no single migration path to the next generation of enterprise communications and collaboration systems and services, and Enterprise Connect delivers what you need to evaluate all the options. Register today and learn about the full range of platforms, services, and applications that comprise modern communications and collaboration systems. Register with code MPIWK and save $200 on the entire event and Tuesday-Thursday conference passes, or for a Free Expo pass. It happens in Orlando, Fla., March 17-19.
The Agile ArchiveWhen it comes to managing data, donít look at backup and archiving systems as burdens and cost centers. A well-designed archive can enhance data protection and restores, ease search and e-discovery efforts, and save money by intelligently moving data from expensive primary storage systems.
2014 Analytics, BI, and Information Management SurveyITís tried for years to simplify data analytics and business intelligence efforts. Have visual analysis tools and Hadoop and NoSQL databases helped? Respondents to our 2014 InformationWeek Analytics, Business Intelligence, and Information Management Survey have a mixed outlook.
InformationWeek Tech Digest, Nov. 10, 2014Just 30% of respondents to our new survey say their companies are very or extremely effective at identifying critical data and analyzing it to make decisions, down from 42% in 2013. What gives?