After battling a lackluster economy for years, most executives are out of ideas for increasing revenue and lowering operating costs. But the smart execs are reviewing history to predict which customers will buy more or splurge on high-end products and enticing them with strategic advertisements. And some are analyzing data to hone inventory purchases or decide how to optimally deploy resources.
The savvy executives don’t have a crystal ball, but they have invested in software and staff to conduct data mining and predictive analytics. Research from Accenture confirms that high-performance businesses are five times more likely to use analytics strategically when compared with low performers.
“Business leaders can avoid mistakes and predict future demand for products and services by utilizing data mining and predictive analytics,” says Dr. Zinovy Radovilsky, professor of management for the College of Business and Economics at California State University, East Bay. “Unfortunately, most don’t know where to start, so they continue to make decisions based on managerial opinion instead of facts.”
Smart Business spoke with Radovilsky about the opportunities to control costs and increase revenue through mining and predictive analytics.
What are analytics and data mining?
Analytics is a diverse field of statistical, qualitative methods and models used for predicting future business trends and customer behavior, and savvy executives are using the practice to make opportunistic business decisions. The process starts when professionals extract or mine data so it can be analyzed and used to identify relationships between the predicted parameters and other factors. The analysis phase is called descriptive analytics, which helps organizations discover what happened in the past, why it happened and how these events impacted the business. Predictive analytics uses the results of both data mining and descriptive analytics to make predictions and optimize business decisions.
How can mining and analysis turn data into dollars?
Analytics may highlight ways to increase customer retention or cross-sell certain additional products and services. At the same time, predictive analytics can reduce operating costs by predicting demand so companies can better forecast inventory or reduce wasted resources. The need for predictive analytics is spreading across various industries and business functions, like marketing, finance, operations, supply chain and human resources.
Predictive analytics and data mining help executives forecast future demand by analyzing customer behavior and profitability by market segments, so they can boost revenue and profit margins by selling additional high-value products and services to certain customers. For example, 1-800-FLOWERS.com attracted 20 million new customers and increased repeat business 10 percent by employing a real-time decision manager that uses predictive analytics applications, business logic and historical purchasing data to motivate customers by offering flower arrangements that appeal to their personal preferences.
How can predictive analytics and data mining reduce operating costs?
These examples illustrate how data mining and analytics can reduce operating costs.
- Predicting future demand helps operations and supply chain managers develop accurate inventory forecasts, purchase the right amount of supplies and eliminate unnecessary waste.
- Predictive analytics can substantially improve allocation of critical resources including equipment, labor and material. For example, after developing and implementing an optimization model to allocate small boat resources, the U.S. Coast Guard reduced its small boat fleet by some 20 percent and overall fleet operating cost by around 5 percent.
- Response modeling allows companies to identify repeat customers from the outset of the relationship and reduce the cost of mailing or calling by targeting only those who are likely to respond. The bottom line is that companies can cut marketing costs by targeting fewer customers while getting the same response.
How does a lack of data analysis or an inability to forecast future events restrict a company’s success?
Companies have accumulated a substantial amount of quantitative business data about their products and services, customers and suppliers. But, their ability to create a competitive advantage by utilizing the data and employing appropriate quantitative models varies significantly. In fact, two-thirds of U.S. companies surveyed by Accenture acknowledged that they need to improve their analytical capabilities. Some organizations still rely on executive opinion instead of using data and analysis to predict the future and make prudent business decisions.
What should executives consider before embarking on a data-mining mission or investing in software or experienced personnel?
Implementation of data mining and predictive analytics can be very time consuming and requires changing the existing decision-making processes and culture, hiring analytical staff and making investments in computer technology. Executives should consider several important things before implementing and managing predictive analytics projects.
First, clearly formulate strategic goals of using predictive analytics and data mining, and identify where the tools will likely make a difference. Then, prioritize the goals by their business impact and ease of implementation and utilization. Finally, identify prospective return on investments before purchasing software, hiring staff and training current managers to use analytics.
Dr. Zinovy Radovilsky is a professor of management for the College of Business and Economics at California State University, East Bay. Reach him at (510) 885-3302 or [email protected]