The Changing Face of Business Intelligence

Howard Dresner originated the term business intelligence. In theearly 1990s, he defined businenss intelligence (BI) as “a set ofconcepts and methodologies to improve decision making in businessthrough use of facts and fact-based systems.” While many of us focusedon data integration and believed that data warehousing wasleading-edge, Dresner created the vision that shaped what we know asbusiness intelligence today.

The business intelligence that was once visionary is nowcommonplace, but sometimes disappointing. Tomorrow’s businessintelligence must become something very different. Too much of today’s business analytics has little connection with real business analysis. At times I am tempted to declare that “the emperor has no clothes.”

But I believe that a significant BI shift is about to occur.Conditions are aligned to drive change. Economic factors demand smartbusiness. New expectations for corporate and executive accountabilityraise the stakes. Consolidation of the BI tools market opens the doorto new and innovative vendors. The next evolution of BI will happensoon, it will happen quickly, and it will expose and overcome theself-delusion that is part of business analytics today.

In recent years, we have strayed from Dresner’s early vision.Current definitions describe business intelligence largely as tools andtechnology. Some fail to mention businessand others include it almost as an afterthought. The next evolution ofBI must return to the vision, enrich that vision and expand upon it tocreate opportunity for truly intelligent business. The nextdevelopments in business intelligence will occur in five significantareas:

• Compelling definition
  
  
• Focus on business analytics
  
  
• Closing the gap between analytics and analysis
  
  
• Focus on business analysts
  
  
• Focus on business

Compelling Definition

The evolution of business intelligence begins with a definitionalshift. Perhaps the most widely quoted BI definition today is DavidLoshin’s “the processes, technologies and tools needed to turn datainto information, information into knowledge, and knowledge into plansthat drive profitable business actions.” Larissa Moss describes BI as“an architecture and a collection of integrated operational as well asdecision-support applications and databases that provide the businesscommunity easy access to business data.” And Steve Dine defines it as“the process, architecture, technologies and tools that help companiestransform their data into accurate, actionable and timely informationand disseminate that information across the organization.”

While all of the definitions are technically correct, it feels likesomething is missing. The troubling thing is that all of thedefinitions are IT-centric. They describe processes, tools,technologies, data, databases and applications. In each of the fewinstances where the word “business” appears, it is used as an adjectivethat qualifies seemingly more important nouns: action, community anddata.

So what is compelling about these definitions? What do they offeras motivation for a business to spend time, money and energy onbusiness intelligence? What do they provide to a BI program as purpose,direction, and the basis for goals and measures of success? I thinkthat they fall short on all counts.

I define business intelligence as “the ability of an organizationor business to reason, plan, predict, solve problems, think abstractly,comprehend, innovate and learn in ways that increase organizationalknowledge, inform decision processes, enable effective actions, andhelp to establish and achieve business goals.” This definition, Ibelieve, is compelling. It describes the qualities of an intelligentbusiness and is sufficiently specific to serve as the basis forpurpose, direction, goals and measures. Equally important, it reflectsand builds upon Dresner’s BI vision. To further understand the originand the implications of this definition, see my April 2008 BeyeNETWORKarticle BusinessAnalytics – Getting the Point.

Focus on Business Analytics

The second major transformation in the changing face of BI is ashift of attention from data to analytics. Don’t be tempted to read“focus on analytics” as “focus on dashboards and scorecards.”Dashboards and scorecards are not analytics; they are simply usefulways to deliver metrics to analytic processes.

Business analytics encompasses the science, disciplines andprocesses of business analysis. It follows reporting (which may usedashboards and scorecards) and precedes understanding (which is done bypeople). This is the point at which information leads to knowledge – ittakes both analysis and understanding to achieve knowledge. To achieveuseful knowledge, the analysis and understanding must have purpose. Forbusiness analytics, that purpose is positioning to reason, plan,predict, solve problems, innovate and learn – the defining capacitiesof intelligent business. Figure 1 illustrates the role and placement ofanalytics in business intelligence.

The point of business analytics is knowledge: knowing what hashappened, knowing why it happened, knowing what to expect in the futureand knowing what to do about it. This is the next “hot spot” ofbusiness intelligence. There is power in analytics, but alsocomplexity. It involves statistics, profiling and pattern recognition,behavioral analysis, time series analysis, predictive modeling,visualization, cause-and-effect studies and more.

The complexity that makes business analytics powerful also presentsa dilemma. Despite integrated data and powerful tools, most businessanalysis is performed by loading local data into simple spreadsheets.Anecdotal evidence suggests that as much as eighty-five percent ofbusiness analysis is actually performed using "manualytics" processes. It is clear that a gap exists between the potential of business analytics and the realities of business analysis.

Closing the Gap

The analytic gap shows itself in virtually every organization astwo distinct and highly polarized approaches to business analysis.Figure 2 illustrates the polarity as described by Gartner. TheIT-intensive approach is one of managed reporting, end-user query,OLAP, dashboards, scorecards and data mining. It is regarded asexpensive, rigid, slow, inaccessible, server-centric and dependent upona large infrastructure. The office productivity approach is dominatedby Excel spreadsheets and Access databases. It is a world of datadumps, locally created data, manualytics and spreadmarts. Thisdesktop-centric approach is considered to be non-scalable, untraceable,unrepeatable, unsecured and particularly difficult to audit.

Despite negative characterizations – expensive, rigid, untraceable,difficult, etc. – both of today’s analytic approaches have merit. Thegoal of next generation analytics is not to choose – certainly not toeliminate one approach in favor of the other. Instead, it is to fill inthe middle, moving from two extremes to a continuum of analyticoptions. Trends in BI practices and in analytic technologies showmovement in the right direction.

Among the practices that will shape the future of business analytics are:

Pervasive BI– Extending the reach of business intelligence into the businesscommunity by reaching more people at all levels with valuableinformation. The key to pervasive is more people.

BI for the Masses– Extending the range of BI capabilities that are within the grasp ofsmall businesses and small IT departments. Where the pervasive focus ismore people, the goal here is more affordable.

Role-Based BI– Analytic outputs are tailored to the needs and interests of specificaudiences. Measures, metrics, trends, scorecards, dashboards, etc. aretailored to business functions (research, marketing, sales, finance,etc.) and to business level (strategic, tactical, operational).Role-based business intelligence seeks to achieve more relevance.

Discovery-Based Analytics– Interactive, exploratory, investigative analytic processes recognizethe natural cycle of analysis where each answer brings new questions.Discovery analytics enable the principle of “listening to the data” tolearn what it can tell you. The goal here is more knowledge.

Agile Analytics– Providing rapid response to situations where immediate need foranalytics exists. Agile analytics encompasses the ability to quicklyand continuously adapt to changing circumstances, both business andtechnical. The agile objectives are more adaptable and faster.

None of these trends, of course, are practical without the aid ofsupporting technology. Innovative companies now offer tools andtechnology to enable the next generation of analytics. Some of theinteresting products are:

Cloud9 Analytics – role-based, on-demand business intelligence applications using asoftware-as-a-service (SaaS) model.

eThority – web and desktop based, user-focused, accessible and scalable approach to business-drivenanalysis of enterprise data.

illuminate – a correlation database that brings agility to the back-end data integration tasksthat are barriers to agile analytics.

Lyza – depolarizing and “filling the middle” with desktop-based data gathering, dataanalysis, reporting and analytic publishing.

Netmap – a visual approach to discovery-based analysis.

PolyVista – extending OLAP with prepackaged, easy-to-use data mining and discovery automation capabilities.

QlikView – rapid deployment of visual analytics from back-end data integration to front-end dataviews.

This is certainly not an exhaustive list, but a sample of the kindsof products that are shaping the future of analytics and changing theface of business intelligence. Each company in different wayscontributes to closing the analytics-to-analysis gap.

Focus on Business Analysts

Analytics are an important aspect of business measurement andperformance management. The analysts, however, are even more importantthan the analytics. It is analysts – the people who perform analysis –who find meaning in the data. These are the people who explorecause-effect relationships and who guide decision-making processes. Itis they who will lead the charge to reshape decision making inbusiness.

The shift to analyst focus is already underway. It goeshand-in-hand with the focus on business analytics. The goals oftechnology providers – user focused, ease of use, desktop based, agile,visual and accessible – all recognize and respond to the important roleof business analysts.

But it takes more than technology. To achieve the right focus we must first answer the question: Where are the BusinessAnalysts?Bear in mind that analyst focus is not reserved for those with businessanalyst job titles. Every manager in a business is a de facto businessanalyst. The controller performing cash flow analysis, the complianceofficer performing risk analysis and the marketing manager analyzingcampaign effectiveness all have some business analyst roles andresponsibilities. These people are analytic professionals, though theymay not be professional analysts.

Focus on Business

The final piece in the BI evolution puzzle is focus on business.The concentration here needs to be much deeper than the lip-service tobusiness alignment. BI and business need to be consciously and activelyaligned in three dimensions: management, motivation and measurement.Figure 3 illustrates this multidimensional view of analytic alignment –a business-oriented BI framework.

The management dimension is used to achieve functional alignment. It describes whatis managed and measured in analytic systems – the functional domainsthat are areas of management responsibility. The diagram in Figure 3shows eight domains that are common to virtually every business. Don’thesitate to adapt and customize these to be specific to your business.Those in the insurance industry, for example, might choose to showclaims, actuarial and underwriting as items in the managementdimension. Higher education may show education, research and studentservices. Retail might include merchandising, customer relations,supplier relations, etc.

The motivation dimension supports goal alignment. It describes whywe measure and manage – the criteria used to determine quality ofmanagement. The diagram illustrates four criteria: performance,compliance, profit and risk. This dimension may need to be tailored tothe nature of your enterprise. Public sector organizations, forexample, may need to include public service and public perception.Higher education institutions will certainly want to includeaccreditation.

The measurement dimension connects management and motivation with analytics. It describes the howof measurement-based management. The framework shows six elements thatapply to enterprises of all types and in virtually every industry. Ameasure is a single, quantitative data value coupled with datadescribing the thing that is quantified and the time of measurement. Ametric is a system of measures with sufficient context to provideinformation through sorting, grouping, filtering, summarization, etc.References are the comparative values that give meaning to metrics –the basis by which metric values can be evaluated as “good” or “bad.”References include thresholds, targets, previous values, etc. A trendis a specific kind of reference in which a series of metric values iscompared to observe behavior over time. Indicators are metrics used toevaluate performance against tactical and operational goals. An indexis a composite of multiple indicators that is used to evaluateperformance against strategic goals.

The three-dimensional approach to analytics is a powerful alignmenttool. As illustrated in Figure 3, the framework contains 192 cells.Each cell represents analytic opportunities. When used to align,prioritize and identify analytic needs, the framework places the rightemphasis on the business part of business intelligence.

In Conclusion

Over the coming several months, business intelligence willexperience change that will have broad, deep and lasting impact.Changing focus to simultaneously concentrate on business analytics,business analysts and business itself is significant. Ultimately, itwill change the way that we think about business and the way thatbusiness decisions are made. When thoughtful analysis replaces gutfeel, conventional wisdom, tribal knowledge and “the way we’ve alwaysdone it,” then we will realize the true potential of business analyticsand enter into the next generation of business intelligence. We willtruly enable business capacity to reason, plan, predict, solveproblems, think abstractly, comprehend, innovate and learn. We willfinally come full circle to realize Howard Dresner’s BI vision.