Business intelligence (BI) today is undergoing a transformation. Historically, BI solutions helped organizations with several key issues, such as routine reporting, alerting and monitoring. However, all of these tasks were based on massive quantities of data stored in corporate data warehouses that were traditionally unavailable to business users. This model prospered in a time when business users were shielded from their operational data, and control of that data remained in the capable hands of IT. Although these methods were sufficient in the past when the needs for information were predictable and routine, the needs of todays employees have changed drastically.
Today, business users are data savvy and need immediate access to detailed data and information. They generally know what types of data they need to meet their needs, but the glut of information available in most businesses makes finding and utilizing all the information difficult. Compounding this issue is the fact that the average knowledge worker does not typically have the technical background and skills to sift deeply through raw data warehouse tables and views.
Business users and organizations need the ability to quickly identify issues, causes and opportunities for improvement. Once these issues are identified, they need to be monitored and distributed often to others. With traditional BI, creating and maintaining effective and understandable analyses can take months and require expensive resources to maintain. And by the time they are created, the opportunity is likely lost or a new situation has arisen.
In response, a new BI term has emerged: operational BI. By extending the ability to effectively visualize and analyze data outside of the IT department and throughout entire organizations, operational BI promises to empower day-to-day decision-makers with better analysis around sales and marketing trends, customer interactions, manufacturing plans, inventories and other areas of business. Fundamentally different than traditional BI, operational BI is more intuitive, scalable, visually compelling and accessible to employees at any level of an organization.
When we look at the business needs for operational BI, seven major requirements for effectiveness emerge:
Leverage existing data stores and infrastructure. Traditional BI requires complex deployments where data is extracted, transformed and loaded into yet another format where security may become an issue. It places a burden on IT to install and maintain these repetitive systems and data stores. Todays BI should leverage the existing investment and access data directly. There should be little for IT to install or maintain - ideally, there would be no new databases to install or configure, no new middle tier, no data modeling exercises, no extract, transform and load (ETL), no extensive training classes and no new certifications for IT to achieve. Most importantly, it should adhere to the existing security and authentication models and not require new security measures to ensure compliance. Existing infrastructure may need to be extended in order to support this new kind of user and ensure faster delivery and integration of BI capabilities into operational processes; ideally, the system modifications required by operational BI are minor.
Access data that IT doesnt have. Because each employees analytical needs are different, not all of the data needed for every analysis will be in the data warehouse, regardless of size. In order to be effective, operational BI needs to recognize this and offer solutions for accessing and analyzing that external data. Next-generation BI solutions should connect to and read the usual operational data stores but also be able to acquire text files, Excel files, ad hoc data files, etc. without having to reformat the data or move it into the data warehouse.
Enable the business user to create new reports and views without IT. Generating the exact views and reports that each employee needs to analyze their unique business situation requires far more resources than any IT department has and creates an IT bottleneck for new reports. And even if the resources were available, most business users cannot anticipate every possible view or report they will need going forward. So, operational BI must enable the user to define new views and reports for themselves. In other words, operational BI must support an element of self-service that is not possible with traditional solutions.
Provide the ability to rapidly analyze and problem solve - not just report. Fundamentally, traditional BI is about distributing answers to yesterdays questions. Or, at a minimum, it provides basic access to information that is only the start of the answer and does not allow users to easily see the bigger picture. Traditional BI interfaces are not built for problem solving, as they will typically answer one question but rarely answer the next. Operational BI must allow the user to rapidly explore and analyze their data by asking not only the question at hand, but also the next question, the one after that and so on - essentially, operational BI must be able to answer every possible follow-up question, and do it quickly and easily. Todays solutions must be able to query, summarize, cross-tab, hypothesize, visualize and report on the fly, depending on where the analysis leads them.
Help the business user see the answer faster. Traditional BI tools produce volumes of text-based reports, static charts and dashboards. In fact, dashboards themselves have become the panacea for many of the ills of BI output. But as Stephen Few writes, Most dashboards that are used in businesses today fail. At best, they deliver only a fraction of the insight that is needed to monitor the business.
Few goes on to say: The root of the problem is not technology - at least not primarily - but poor data presentation. To serve their purpose and fulfill their potential, dashboards must display a dense array of information in a small amount of space in a manner that communicates clearly and immediately. This requires design that taps into and leverages the power of visual perception and the human brain to sense and process several chunks of information rapidly.1
Dashboards, static charts and text-based reports can give a glimpse of an answer but often are a dead-end for real understanding. They are often poorly designed and therefore obscure the real meaning, or they dont communicate deeper answers as to why a particular pattern or trend is happening. Operational BI must provide a better and faster way to see patterns and trends - state-of-the-art data visualization techniques and good design principles must be integrated to help business users capture meaning faster and get valuable answers sooner.
Deliver a means for sharing and updating analyses and findings in real time. Traditionally, BI has excelled at generating automated reports that could be distributed across an organization. In fact, one could argue that automated report creation and distribution were the major benefits of traditional BI. But todays operational BI needs to go three steps further. First, not only must it create and distribute automated reports, it must also allow business users to define the permissions of those reports. This means they must be able to not only design and set up the report themselves but also be able to easily configure which recipients can see which views of which underlying data. They then need to be able to publish the reports and views where recipients (with the proper permissions) can access them. Second, it must allow the recipients to conduct lightweight analysis against those reports. The intended audience often has additional questions that should be able to be answered without having to rerun the entire analysis or involve the publishing analyst. And third, these analyses must update in real time in order to make sure everyone is getting the right answers to the right questions each time they look. Real-time updates also eliminate the need to constantly regenerate analyses.
Require no training to start and only minimal training of sophisticated business users. Traditional BI often requires multiday, in-person training courses, even for business users who would only be light users of the BI system. In todays world, this isnt just unrealistic; its a guarantee that only a fraction of potential BI users will ever get any value. So, operational BI must be easy enough to use that nearly any business user can conduct basic analytical inquiries without any training. Naturally, for people who want to learn more sophisticated functions, training should be available via means that are easily accessible: online training courses, on-demand training courses, tutorials, etc., and not just high-cost, in-person classes.
In most of todays business environments, one thing is very clear: People want the ability to ask questions of their data and get answers in real-time, and they are disappointed in the status quo. Analysis is a crucial part of any business and any job title, from an HR representative at a Fortune 500 company to a sales rep at a small startup. In this day and age of endless information, its crucial that businesses take the necessary steps to update or deploy BI solutions that will be accessible, useful, effective and easy to understand for the entire organization. Operational BI promises to remove the roadblocks to effective visualization that traditional BI created, offering a more effective, pervasive and valuable alternative.
- Stephen Few. Why Most Dashboards Fail. PerceptualEdge.com. March, 2007.
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
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access