The Nobel Prize for Economics was awarded to George Akerloff, Michael Spence and Joseph Stiglitz for their pioneering work in the area of information economics in 2001. The premise of their research was that many markets are imperfect because information asymmetries exist in which one party to a transaction has more information than the other party. The subtext here is that information has value and that benefits can accrue to those who possess it.

The views of Akerloff, Spence and Stiglitz are simple but profound in that they challenged the fundamental economic assumption that markets had perfect information, i.e., that everyone had pertinent information.

Today, we find a striking parallel to the world of IT and business intelligence (BI) in particular. BI initiatives are often based on the false assumption that users know exactly what information they want and need. More specifically, we have observed that BI initiatives typically focus on existing information rather than on the set of information that truly influences decisions and their outcomes.

This latter set of information can have the most powerful impact on performance. Whether IT is embarking on or has completed a BI implementation, defining this information will help unlock the key to more effective decisions and give organizations a significant advantage.

Why BI?

Identifying BI information and analysis requirements depends largely on how one defines the purpose of BI and what organizational goals it seeks to support.

In the most general sense, the goals of an organization are fourfold and reflect two different time horizons - current and future. Goals are:

  • To perform processes in the execution of the current strategy,
  • To comply with current laws and policies,
  • To identify future opportunities, and
  • To manage future threats and risks.

The purpose of BI should be to provide information that supports decisions behind all four goals. Viewed another way, BI should support decisions that not only help execute the current strategy but also validate whether the strategy itself is optimal and point to areas for refinement.
This information goes well beyond accounting and customer relationship management (CRM) data to include multiple perspectives:

  • Economic intelligence is information about the organization's financial performance and value creation. It can also include external information, such as trends in household disposable income and raw material costs (e.g., oil).
  • Marketplace intelligence is information about the organization's customers, distribution channels, competitors and substitutes. This can include such information as customer behavior, market share and competitor initiatives.
  • Operational intelligence is information about the organization's operational effectiveness (including process cycle time, cost and quality) and risks (such as the potential for supply disruptions).
  • Organizational intelligence is the collective set of knowledge, capabilities and experience held by employees.

The value of this information can be significant. At the consumer level, Amazon is a wonderful example. Amazon offers customers valuable information for making decisions on which books to buy. When we search for a book, Amazon gives us much more than the book's cover, price and availability. It provides user reviews and suggests other books, all of which inform our decision. Imagine if users within an organization had the same depth and relevance of information that Amazon gives its customers.
A mortgage client of ours provides an instructive corporate example. The client collected extensive information on customer transactions and Web site clicks. However, it was not able to understand why its offerings were not more widely accepted. Looking at the problem from a BI perspective revealed the need for greater insight on customer segments and behaviors, which ultimately helped it to achieve a multifold increase in originations and to identify new product opportunities.

What Do You Need?

BI implementations typically gather information and analysis requirements by asking users, "What do you need?" We have found that users typically assume they have the information they need and instead focus on two types of requirements:

  • Automating reports that are currently produced manually, and
  • Creating drill downs or multiple views of information.

As an example, the accounting department may want a BI system to automate a manually created report for allocating expenses, or the sales department may want a BI system to view sales for different types of accounts.
Both requirements illustrate a technology-versus-information mind-set.

Why do users not know or reveal all of their information needs? There are clearly many reasons, but three of the most common ones we've observed include the following:

  • Focus is on technology. BI projects often arise as a technology or database initiative. This origin sets a tone for how technology can be used to view and analyze information. Also, the BI projects are typically led by IT and technology vendors, who often assume that the business users know what they need.
  • Answers are in the drill down. There is often an instinctive assumption that a greater understanding of performance comes from having more detailed or different views of the information they already have. For example, when the sales department is not meeting goals, users frequently want breakdowns of sales by customers, products, regions and myriad other views, assuming this information will quickly reveal the underlying problem. This information is useful for identifying where a particular problem may be but will not provide insight into why the problem exists.
  • Users rely on intuition. Typical corporate systems - e.g., accounting, CRM and manufacturing systems - contain information that is inherently quantitative and relatively easy to define and report. Accounting systems, for example, are governed by more quanitative information - e.g., generally accepted accounting principles (GAAP). However, there is no comparable GAAP around the factors that influence performance and decisions - such as customer behavior and product quality. This information can be highly qualitative and require significant time and resources to collect and define, which organizations are reluctant to provide. As a result, individuals will rely on intuition and conventional wisdom to guide them.

There has been a greater push toward using balanced scorecards, which incorporate qualitative information. Scorecards, however, are not sufficient: they have yet to be used beyond the executive suite in any significant way. They often lack depth around what is driving performance, and they focus on measuring strategy execution rather than whether or not the strategy itself is correct.

Recognizing the Symptoms

How do you know if your organization is not collecting the right kind of information? From our experience, some of the more prominent warning signs are as follows.

Users keep asking for more reports. Users will hope that analyzing the same information in different ways will lead them to the answer. This rarely works well and points to a need to re-assess fundamental drivers behind performance instead.

The IT group at one client suspected this issue when users began severely straining IT resources with more and more report requests. All the while, profits continued to decline. Before investing more in BI technology, the client asked us to look at their true business drivers and information needs. We came up with powerful findings: only a third of the information they needed was collected and distributed, a third of the information they needed was not collected, and a third of the information they needed was collected but not accessible by users. The quality of information - not technology - was clearly the main issue.

Strategy keeps changing. When an organization's strategy doesn't seem to be working, certain managers quickly change direction. Most of the time, such rapid changes are a poor surrogate for understanding true business drivers and collecting related information. At some point, customers will abandon the company.

This situation was prominent at one of our other clients. When we determined what information was relevant to collect, particularly around the area of customer behavior and competitor offerings, the client was able to understand where to focus and quickly realized increased revenue.

Decision-makers rely heavily on anecdotes instead of facts. Information is not as easy to collect as the typical transaction, such as sales and expenses. In the absence of readily available information, there is a bias to rely on personal experience or anecdotes.

Identifying the Right Information

IT groups include terrific technologists but usually lack the strategic skill set to identify many critical information needs. Likewise, users often have deep process or functional skills rather than strategic skills. So how do BI teams - with IT and/or business line members - identify information that truly provides value to users?

To solve this challenge, the best course is to engage a strategic resource or adviser who can excavate and translate user needs. The BI team, along with this strategic adviser, should conduct the following critical steps:

  • Understand overall objectives. Overall business objectives are the starting point for any initiative. However, objectives are often implied rather than stated and can vary from person to person. As a result, it is essential to articulate and achieve consensus on them.
  • Identify key decisions for each major process and role. Ask individuals what primary decisions they make. Decisions will be either strategic, tactical or operational by nature. Since information is the basis of these decisions, it is critical to catalog them.
  • Ask why and how. Understanding why and how users make the decisions they do is just as critical as the decision itself. Answers to these questions will reveal a wealth of information about what drives a good decision. Often, individuals' responses will include information that is not captured but should be.
  • Derive information needs. Once the above-mentioned steps have been completed, it is possible to derive what the important information needs are. One will find that many critical needs are not collected or accessible.

 Not all critical information needs can be embedded into a structured BI database for later reporting and analysis. Certain information may be too expensive to collect or too unstructured. In the latter case, other tools, such as portals, may be most effective. These are judgment calls that will have to be made. Nonetheless, focusing the lens more closely on true information needs will enable IT and BI teams to contribute to more certain decisions, more effective strategy execution, greater opportunities and, in turn, much better performance. 

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access