Building an analytics-driven organization

As the pace of business continues to accelerate, the need for fact-based, analytics-driven decision making is becoming more and more critical to success. Across all industries—from banking and insurance, to telecommunications, to consumer goods manufacturers—the organizations that consistently outperform their competitors have placed becoming an analytics and data-driven business at the core of their enterprise.

While the demand for advanced analytics and data management has never been greater, many organizations struggle with making this transformation and often end up with a disconnected set of analytic models, data warehouses, reporting tools, and consequently, poor results.  The key to success is building a sustainable framework that combines advanced analytics with proactive business practices. 

We have found the following advanced analytics framework to be a simple, effective approach for building an analytics-driven organization.


Let’s look at each of these steps in a little more detail.

The first step in implementing an advanced analytics framework is to clearly define the business problem or challenge you’re addressing.  Are you trying to reduce customer loss or attrition, increase profitability or lifetime value through cross-selling, or decrease costs by reducing or eliminating unproductive processes?  Can you define the problem through relevant metrics and data?  Can you measure it across customers or time periods?

Effectively defining the problem you’re trying to solve is critical to determining the data you’ll need, how to prepare and manage it, the analysis to be performed, and finally how you’ll deploy the resulting insights across your business.

Once you’ve defined your problem, you need to analyze it deeply and identify its root causes. These root causes are the key drivers that best explain the behavior in question.  Done properly, this root cause analysis can quickly find patterns in your underlying data and turn those patterns into rules and models—meaningful business insights and rules that you can then turn into concrete strategies and tactics. 

This step requires advanced data mining capabilities that enable you to quickly and accurately identify root causes. A tool such as Hypercube®—a sophisticated data mining tool developed by our global alliance partner, BearingPoint—can enable this level of analysis. Hypercube® utilizes a unique, proprietary, and non-statistical technique to exhaustively analyze every possible combination of your data to explain and predict business outcomes. Its analyses produce root-cause findings that you can use to modify business processes, define business rules that serve as alerts or triggers, and/or provide input into predictive models or algorithms in the next (“Predict”) phase of the framework.

Once you’ve diagnosed your business challenge, the next step is to use these insights to build a predictive model or algorithm that can be applied to your customers, interactions, or business processes.  At this point, the root causes identified in step two are used in combination with predictive modeling techniques to forecast what’s likely to occur in the future.  The resulting model provides the ability to score customers or processes in real-time for proactive strategies and optimized decision-making. 

This step is critical in building a comprehensive analytics framework, since the root causes found in the diagnosis step may only occur occasionally, but with a high degree of certainty in the outcome.  A predictive model complements these findings for the remainder of cases where the outcome is less certain. 

Turning insight into foresight will require strong data mining capabilities, such as those inherent in Hypercube® described above, as well as advanced statistical techniques such as decision tree analyses, multi-variate regression (linear and logistic), sophisticated time-series forecasting, and simulation models.

Prevent or Promote.
The final step in this framework is often the most difficult, since it involves new processes, change management, and stakeholder adoption.  The goal here is to put the analytics into action by designing or re-designing business processes based on the root-cause findings, and executing tactics based on your predictive model scores. 

If your goal is to prevent customer attrition, for example, you’ll need to correct the conditions that were found to be the most likely cause of defection and then execute proactive retention activities to the most at-risk customers based on their attrition scores.  Conversely, if increasing lifetime value through cross-selling products is your objective, then you’ll want to replicate the conditions found in the root-cause analysis and target your cross-sell promotions to customers with the highest propensity for your additional products.

The critical elements: flexibility and discipline.
For any framework to be effective, it needs to be flexible enough to accommodate a wide array of business needs, from solving customer experience or marketing challenges to improving internally-focused operational processes.  The analytics framework described above has proven effective in a variety of industry and business circumstances—from enabling insurance companies to better manage high-risk claims, to helping pharmaceutical companies maximize patient adherence to prescription regimens, to equipping retail banks to maximize customer lifetime value and loyalty.  The key is a disciplined approach to identifying root causes, making accurate predictions, and developing insight-driven strategic processes and programs for taking action and addressing your business challenges.

For more information about West Monroe Partners’ advanced analytics solutions or to discuss how we can help you establish analytics capabilities that support your business, please contact us.