Leveraging Business Intelligence - Making Your Organization More Analytic

Your company has invested large sums in buying a new ERP or CRM system or in building a new data warehouse. Now your CEO would like to see your organization use this data to make better decisions. Tom Davenport and his colleagues wrote a recent article that provides some valuable insights into this issue (Davenport, Harris, De Long and Jacobson, "Data to Knowledge to Results: Building and Analytic Capability," California Management Review, Volume 43, Number 2, Winter 2001, pp. 117-138). In this short review article, I will draw heavily on the concepts outlined in the article and supplement the authors' insights with Knowledge Management Associates' experience in designing and implementing OLAP systems, to examine how an organization can make better use of the vast amount of transaction processing system data it has collected.

Davenport and his colleagues suggest three steps toward building a capability within your organization:

  1. Set the context—a foundation of technology, strategy, expertise and organization culture must be in place before you can begin.
  2. Transform the process—change both the processes used to analyze data and to make business decisions to take advantage of these new data sources.
  3. Measure the results—provide feedback to the analysts and managers so they understand the impact of the decisions they have made.

Set the Context

There are four key contextual factors that must be in place:

  1. Strategy—Your analytic resources must be focused on problems that matter to the success of the business. Start with the organization's strategy. Is your CEO trying to close more large sales? Focus your analytic resources on tracking these sales, measuring the sales pipeline and learning why you are making and losing specific sales.
  2. Technology—Most organizations have collected the data they need. Technology managers, however, must often spend considerable time and effort to clean the data and organize it. This up-front work is critical to the success of your organization's OLAP endeavor.

    The technology to build data cubes (e.g., sales by product, by region, by customer over time) and to move data into cubes is available from many vendors. For example, you can use Microsoft's SQL Server OLAP Services, Excel, PowerPoint, and Digital Dashboard to extract data, analyze it and present it. These are tools that are powerful, easy-to-use and familiar to many technology experts and business analysts. Furthermore, they are often already present in your organization.
  3. Skills—The skills needed to analyze large quantities of data and using this analysis to make better decisions is not found in most organizations. Davenport and his colleagues identified several skills you will need:
    • The technology skills to understand the workings of your organization's transaction processing systems, so that the needed data can be extracted, organized and presented to your business analysts.
    • The analytic skills needed to transform this raw data into business insights. This could range from generating simple charts and tables to summarize the data to building sophisticated models.
    • Sufficient knowledge of the data to understand and explain why the results of a marketing analysis are not consistent with the data presented in last quarter's income statement. This requires deep knowledge of the structure of your organization's data (e.g., how does the aggregate value of orders in the ERP system for a month relate to revenue for the same month?) Davenport states, "This knowledge is tacit, changeable, and idiosyncratic, but without it the value that can be derived from decision making is very limited (p. 125)."
    • Enough understanding of the business itself so that you can ask the right questions and make sense of the data.
    • The communication skills to get your message across to key members of the organization.

  4. Organization culture—Your organization's culture must be fact-based. This has to come from the top. Senior managers must ask the questions (e.g., which of our customers are good candidates for cross-selling?), insure that resources are available to answer the questions, and act on the results of the answers.

Transform the Process

For an organization to generate the business knowledge needed to make better decisions, managers must intertwine the business analysis process with the decision making process. Quite simply, business analysts do some simple data analyses (study sales by product by region), draw some conclusions (there is an opportunity to grow sales of a specific product group in the Western states), and present the results to managers. The managers have comments and questions on the results. The analysts learn from the managers, refine their analyses and repeat the cycle. At the same time, the managers take actions based on the analysis. The results of their actions are then collected and analyzed. These highly inter-related processes lead to more learning on the part of the analysts and the managers, better analysis and better decisions.

These processes evolve over time. Analyses get more sophisticated as analysts' skills and data knowledge increase. Analyses also change as the business environment changes, old problems go away and new ones arrive.

Measure the Results

Leveraging OLAP and increasing your organization's analytic capability is a way to improve business results and not an end in itself. Your organization must measure the results of the transformed business processes through an improved management reporting system.

Many organizations are moving to a balanced scorecard approach to implement financial, customer, process and learning measures. These measures can be organized into a Digital Dashboard and presented on the corporate intranet. This provides managers and analysts with key feedback on the effectiveness of the decisions they have made based on the analyses they have carried out. This measurement system will also evolve over time as the business environment changes and the organization's analytic capability is re-directed to solving new business problems.

How Do I Get Started?

Becoming a more analytic organization might seem like a daunting task that requires significant effort over a long period of time-and it is. For many organizations, however, it is critical to their long-term success. Organization learning is the only sustainable competitive advantage available to most organizations.

The good news is that you can start small, see some significant short-term benefits, and build your organization's analytic capability over time. You don't need to build an expensive data warehouse, hire Ph.D.s in statistics, or buy complex modeling software.

You can start with a pilot. Choose a business issue that is important to senior management. There is a very good chance that there is an existing transaction processing system that collects data on this process. You can build a cube in as little as a few hours, perform some initial analyses, share it with senior management and build some momentum.

You can grow your analytic capability from this modest start as you grow your organization's skills and transform your culture to be more fact-based.