Abstract
We will address how business intelligence value means getting the right information to the right user at the right time; why some business intelligence systems fail to deliver this value; how
organizations can do a better job of achieving this goal; and how business intelligence tools can benefit staff at all levels in the enterprise, including senior and middle management, as well as
frontline employees. Additional highlights include examples of companies that have successfully translated business intelligence into business value.
Earlier this year Gartner, Inc. announced findings from a survey of 1,300 CIOs representing more than $57 billion in IT spending (Gartner EXP, 2005). Their report indicates that business
intelligence (BI) is the number two technology priority among the survey participants. In addition, executives are beginning to push BI applications further into their organizations through what
industry analyst Butler Group described as “pervasive business intelligence” (Butler Group, 2004).
Further evidence of this trend is found in the InfoWorld Business Intelligence Report 2005, which discloses that during the next year, 70 percent of respondents plan to increase the number of
employees who have access to BI solutions. (Gincel, 2005)
Driving this pervasive approach is BI’s evolution. Powerful analytical tools, alerts, and feedback mechanisms can be made available cost-effectively to decision makers at every level
throughout the organization. Because these more “pervasive BI” systems can bring new efficiencies and generate high return on investment, they are creating an environment where,
increasingly, business intelligence equals business value.
New Classes of Users
As an example of how pervasive BI reaches out to new classes of users that span the enterprise, imagine that a financial services company has launched a new product called QuickLoan. Top management
sees overall company year-to-date revenue falling, but the executives are unable to pinpoint the causes and might even entertain erroneous conclusions—such as a drop in new customer
generation—based on the limited data in their current view.
With its pervasive BI system, each functional area sees key performance indicators (KPI’s) relevant to its point of view. Middle management sees a chart enabling executives to determine that
the downward trend is probably related to defects (e.g., improper processing of loans), not lack of new customer generation. This analysis is further supported by another chart indicating a rise in
customer complaint calls, presumably triggered by the defects.
The bottom-level “frontline” staff sees different charts, enabling the employees to target the specific problem— increased defects in the processing of their QuickLoan financial
product that have led to a drop in QuickLoan revenue and increased customer complaints—as a potential cause for the drop in overall revenue. Each layer of management can see these performance
trends, from the bottom up or from the top down, and decisions can be made to address the root of the problem, not just the most evident symptoms.
Because decision makers throughout the organization have data relevant to their area of responsibility and the ability to communicate key performance trends—in near real time or “right
time”—they can make collaborative decisions that ultimately take advantage of opportunities or help deal with adverse conditions.
Challenges to Full Deployment
Pervasive BI fosters the kind of enterprisewide approach illustrated in our example by overcoming some of the challenges that have hampered the full deployment of BI tools throughout the
organization. These challenges include:
Lack of Skills
One way in which organizations have addressed the lack of skills and the complexity of BI tools is by developing competency centers. A competency center is a small group within an organization that
has skills in specific BI tools that executive leadership wants as standards throughout the enterprise. The competency center champions applications that can be distributed to departments and
divisions along with the required training that enables employees to leverage the BI solutions. By standardizing on specific tools, organizations are able to reduce the overall complexity of BI
tools and their implementation.
Cultural Resistance
Organization leadership has been reluctant to allow deeper organizational penetration of BI data for a variety of reasons, including the cost of a broadly distributed system, concerns over adoption
and usability, and problems with the information that might be made available beyond the executive level.
Some experts believe this resistance is going to fade as the value of more broadly disseminated BI programs are proven. Even when executive leadership wants to implement a more extensive BI
approach, resistance is still an issue because employees are faced with changing the way they do their jobs. In setting and communicating an organization’s enhanced BI strategy, managers need
to anticipate and address such issues as resistance (due to fears over losing control and learning new systems) as well as uncertainty (about how to use enhanced access to information from those
who are not accustomed to acting on their own initiative). With BI, as with any technology initiative, the enterprise must be ready to deal with change. This includes effectively communicating to
the organization’s decision makers why the change is worth the effort.
Tools are Unavailable
BI systems often don’t deliver because the tools are too complex for frontline decision makers, or the organization’s leadership resists providing BI tools to these employees. When BI
tools are not available throughout the organization, business people at every level cannot leverage the power of the system to make the best decisions and produce the best results for the business.
Tools Lack Relevance
Some BI systems fail to deliver on their potential because they are not relevant to how many managers operate. Henry Mintzberg, a professor of management studies at McGill University and author of
several books and studies on managerial practice, states, “Managers’ activities are characterized by brevity, variety, and discontinuity… They are strongly oriented to action and
dislike reflective activities.” (Mintzberg, 2003) Managers are conditioned to respond quickly to problems or opportunities within the business. Generally, managers don’t have time to
spend hours digging into a BI application. They usually assign that work to someone else. For a BI application to hold value for a manager, it must be easy to use and provide current data the
manager can act on quickly.
Inadequate Data
Sometimes data in a BI system has little value. For instance, the data might not be timely enough. In some businesses, managers need updated data every hour, but only receive updates weekly. This
can occur with a BI system that is developed for one set of users and then extended to support a second group that has different requirements, when poor requirements gathering or planning occurs,
or simply because business conditions have changed. When hourly updates are required but weekly updates are delivered, the system cannot succeed—it does not provide the desired value to the
target users.
Alternatively, sometimes poor data quality hampers a BI application. Generally this occurs if data isn’t clean enough when it comes into a BI system from the source system(s). The bad data
gets combined with good data and the result is overall poor quality information. Once a system loses credibility with users, rebuilding that trust is a major challenge (it might be easier to
replace the application). Such data failures are particularly damaging because they often are discovered after the system has been developed, deployed, and paid for.
Unclear Results
A BI system can lack the capability to show clear cause and effect. Suppose you have a dashboard that discloses a problem related to specific KPI’s, and you take action to deal with the
issue. If the result of your action doesn’t show up on the dashboard within an appropriate time frame, it’s hard for you to know if what you’re doing is impacting the business in
the way you want. Illustrating cause and effect clearly and in a timely manner in the application is critical to decision makers.
Meeting the Challenges
BI has evolved into its more expansive forms to solve many of the challenges we’ve discussed. For example:
- Where company culture resisted deeper penetration of BI data, executive management now advocate BI throughout the organization because pervasive BI tools are less costly, easier to use, and
have demonstrated a solid return on investment. - Where previously BI data was not timely enough, pervasive BI has evolved to offer reporting and analytical tools that make it easier to get the right information to the right users at the right
time, through multiple levels of the organization. - Where older BI systems lacked the capability to show clear cause and effect, now pervasive BI solutions enable management to promote the organization’s goals by disseminating KPIs and
dashboards so each user and group can see how they are measuring up against those goals in near real time.
Figure 1 depicts one model of how pervasive BI can address the needs of executives at all organizational levels. In a typical hierarchical enterprise, executives occupy the top of the pyramid while
the majority of employees and business partners form the base. The greatest potential audience for a BI solution is the base of the pyramid, and small changes made there will accumulate through the
pyramid into large benefits for the organization. If an enterprise only provides the solution to the top of the pyramid, it’s missing the potential to enhance the capabilities of employees
and managers in the rest of the enterprise.
Figure 1. Persausive BI can potentially address all executive levels
Different Application Interfaces
When delivering a more pervasive, enterprisewide BI system, different types of users will require different application interfaces. Some users may want dashboards with KPIs and scorecards. Others
may want full-featured cube browsers that let them explore the data and discover unexpected insights about the business. Still others might want to perform analysis based on static reports or an
Excel file. Providing wide-ranging data in these different types of interfaces and formats is a high-value activity, and failure to do so only increases the chances that a BI implementation
won’t succeed.
One key to success is to extend the BI solutions to many levels in an organization and allow employees to customize how they use the applications by choosing the most relevant information, setting
thresholds, and specifying notification alerts. In addition, organizations need to ensure the data delivered is actionable by the employees receiving it. Actionable data is data that employees can
use to change their business activities— for example, redistributing the work load because one regional order-processing center is overloaded. Actionable data also is information that
management can use to measure employee job performance.
Another important criteria for success, business requirements, determines how current data in a BI system should be. In a fairly static business, older data will be acceptable. In a business in a
chaotic or rapidly changing environment, the enterprise needs up-to-date data to make good decisions. If the business cannot obtain data in time to meet business requirements, the BI
solution’s value is marginalized.
Today’s pervasive BI systems also help ensure success by providing automated alerts that can be set either by individual users or management, and typically delivered by e-mail.
Finally, a pervasive BI system is designed to allow for change over time. Businesses change, and the systems designed to report on the businesses activities should evolve with them. Making a BI
system a living system is imperative to the long-term success of the project and a key driver of business value.
Integration with Microsoft Office
One development helping make BI more pervasive is BI vendors’ software applications that improve integration between BI tools and Microsoft Office. Until recently, line-of-business workers
faced inefficiencies from poor integration between server-based BI applications and Microsoft Office.
New solutions incorporate powerful data management, analytics, and reporting capabilities into Microsoft Office applications including Word, Excel, and PowerPoint. Custom analyses created by IT
specialists can be made available to business users at all levels of the enterprise from within Microsoft Office.
For example, SAS Add-in for Microsoft Office enables business managers and frontline workers to browse tables and charts via a graphical user interface and execute on demand with the click of a
button. Additionally, business users can refresh the results within a document with a single mouse click. Users can refresh all embedded analyses (which will rerun and return the results using the
latest data) or selectively refresh the results of one or more embedded analyses within the document, all without involving IT specialists.
Additionally, Microsoft recently released a beta version of a new, server-based performance management scorecard application, code-named “Maestro,” that helps users perform deep
analysis by leveraging the Microsoft Office system to build, manage, and use scorecards and KPIs. Microsoft has described Maestro as the next step in the company’s business intelligence
product strategy. Aimed at enterprises that have implemented data warehouses to support reporting and analysis, it enables deployment of scorecards and strategy maps to employees so they can track
and analyze business metrics. The goal is to more effectively align individuals and their areas of responsibility to corporate strategy. (Galli, 2005)
Benefits of the Model
The pervasive BI model clearly articulates management’s goals and priorities by disseminating actionable KPIs to users who can improve the entire organization by making better individual
decisions. Because the KPIs are actionable and measurable, users can clearly see whether they are meeting management’s expectations. Management can also see whether individual users and
groups are meeting expectations.
By implementing this model and including a relatively frequent feedback loop, an organization can operate more efficiently. It can make decisions more rapidly and closer to the “point of
impact,” where problems or opportunities arise.
This model also can help improve corporate governance by providing a process that clearly defines what’s important to management, then determining if individual users or groups are meeting
these goals. This approach provides the enterprise a more accurate view of its business over time.
Realizing the Potential
Innovative organizations are turning business intelligence into business value by implementing pervasive BI systems that accomplish diverse goals, from enhancing efficiency and maximizing
competitiveness to creating new business opportunities and revenue streams. These leading-edge enterprises span all major industries. All see the potential in pervasive BI. For example:
Financial services
Financial services organizations can deploy pervasive BI systems to improve operational efficiency and visibility in back-office operations. Different executives look at data relevant to their
point of view. Senior management sees that overall revenues are down. Operations executives see customer service issues increasing. Frontline sales personnel see that some clients are not getting
their trade confirmations on time and have cut back on their activity. All levels of management can see how these problems are related and can collaborate on fixing them quickly.
Sports
A pervasive BI system can enable sports team owners, coaches, and players to see performance-analysis data appropriate to their needs. An offensive coordinator for a football team can use the
analytical data to create the most effective strategy to use against an opponent in the fourth quarter of an upcoming game. The team’s owner can use the pervasive BI system to analyze
eligible quarterbacks and recruit the player offering the best performance within a specific salary range. Team management can monitor attendance and profitability over time.
Healthcare
Hospitals improve both patient care and administration by leveraging pervasive BI technology. Business executives can pinpoint KPIs such as profitability and revenue projections. Nurses and nurse
managers can track occupancy data (such as average length of stay and discharge information). Doctors can quickly access lab reports and clinical trial information at the point of patient contact
to determine the best course of treatment.
One real-life example of pervasive BI involves a leading global-trade bank with several hundred billion dollars in assets, more than 3,000 branches in over 50 countries, and more than 100,000
employees. The bank needed a more efficient way to monitor and manage its trade processing operations, quality metrics, costs, sales performance, and risk exposure across its global network of
offices.
The challenge was to create a solution that would enable management to quickly gauge and be alerted to inter-day activity (as opposed to month-end static report reviews), and to be empowered to
analyze and address issues proactively. The solution needed to be easy for executives to navigate, yet provide the power to do more in-depth analysis as required. Further, as many executives are
often on the road, information needed to be accessible on the Web. A Phase I solution for the North American region was required within six months.
After an initial requirements analysis, the bank created a business intelligence repository (BIR), a powerful pervasive BI solution comprised of a data warehouse, refreshed on the hour from the
bank’s operational mainframe system, with two different user interfaces. One interface was an easy-to-use, Web-based, point-and-click dashboard for non-technical business management; a
second, more powerful interface was based on a third-party tool for more complex analysis. The BIR solution was developed, populated with data from North America, and fully operational to the
various business areas within three months.
The dashboard is comprised of a set of analysis drillpaths, which are made up of a number of charts and graphs providing simple, one-click drill-down capability. Drill-paths are tailored to the
bank’s business areas, allowing each to view operational information in the context of its business area. Besides security logons, access to drill-paths varies according to each user’s
access rights. The dashboard also provides color-coded chart data (in red and green) that alerts users to possible problem areas and includes drill-down capabilities to quickly pinpoint the source.
The second interface is a third-party tool that enables power users to interrogate the BIR and perform more complex analysis and ad hoc queries. Business value is multifaceted. The BIR helps the
bank’s management track trends, proactively identify problems, balance workload among its global operation centers, minimize risk, and manage costs in its business areas around the world.
Another real-life example of a pervasive BI approach is found in a business information company specializing in industry analysis. The company helps its clients (which include several thousand of
the world’s leading enterprises) address complex strategic issues. Through its proprietary databases and expertise, the company provides its clients with unbiased expert analysis and in-depth
forecasts in a number of industry sectors, including automotive, consumer markets, energy, financial services, pharmaceuticals and health care, technology, and transport and logistics.
The company’s primary business was founded on its research and reporting capabilities, but the company collects much more information than it can disseminate to its customers in print
reports. These print reports provide summarized information and generally do not show the wealth of underlying data the analyst has worked with to arrive at conclusions. By recognizing that within
the more detailed data often lies information that customers might find valuable, the company identified a new market and created a BI system that allows customers themselves to run ad hoc queries
against the detailed data. This data can be retrieved through a Web-based interface, allowing customers to drill down and explore the data valuable to them.
By selling this system as a subscriber service, the business information company has entered into an entirely new market. It has turned potential business intelligence, which the company previously
had been unable to share, into business value for the company and its customers.
Summary
Today’s more scalable, pervasive BI systems take relevant information often only reported to senior management and pushes it down the organization, to management and staff alike. At different
organizational levels, the data is presented so users see only what is most relevant to their daily tasks, broken out into vertical silos, with expectations and performance clearly identified.
When all key decision makers within an organization are using BI tools, BI becomes pervasive, and becomes mission-critical. When this happens, everyone from the CEO to the newest individual
contributor can make better decisions faster, an ongoing goal in today’s highly competitive business environment.
REFERENCES
Butler Group. “The New Business Intelligence,” Butler Group Symposium, October 5-6, 2004.
Galli, Peter. “Microsoft ‘Maestro’ to Tune Up Business Performance,” eWeek, May 10, 2005.
Gartner EXP. “Delivering IT’s Contribution: The 2005 CIO Agenda,” Gartner Executive Programs, January 14, 2005.
Gincel, Richard. “Trickle-Down Business Intelligence,” InfoWorld, April 25, 2005, 47-50.
Mintzberg, Henry. The Manager’s Job: Folklore and Fact, Watertown, MA: Harvard Business School Press, 2003.