Templates for Operational BI Success
Front-line workers, including managers in sales operations, production, finance and logistics, need timely information that is operational in nature in order to make effective day-to-day business decisions. But, too often the data they obtain through an extraction process from packaged application systems is dated or not organized in ways that best help them to make immediate decisions. Operational business intelligence (BI) is a technology solution that helps provide front-line workers with the current, meaningful information that they need at the moment they need it in order to make highly effective business decisions. For those companies considering deploying such technologies, it is important to consider the range of strategic, organizational and technological factors that are critical to ensure successful deployment of a powerful and effective operational BI solution.
Operational BI is the capture and analysis of operational data for the purpose of taking immediate action to improve a business process. This differs from traditional BI, which involves trend analysis for longer-term planning. In the case of financial institutions that are engaged in selling and trading financial instruments, these institutions often seek to provide clients with greater visibility into their account activities to foster greater customer involvement. But most financial applications are isolated and operate in silos, making it difficult for application-specific BI tools to access these disparate systems and present the information in a compelling format. Moreover, end users want a view that combines data from the transaction system such as a trade order management system (TOMS) with data from a data warehouse to capture information for dissimilar securities such as stocks, bonds and derivatives. Furthermore, the information needs to be presented on demand - that is, when the user wants it and in a format that the user understands.
Traditionally, organizations relied on IT to provide separate, operational reporting projects. Individual reports from this process often exceeded 100 pages, and multiple reports were often generated per project - yet, each report would typically be sourced from a separate application using different BI tools and interfaces within a separate architecture or environment. What's more, the reports often were delivered in different formats, some printed and some delivered via email.
The problem with this approach was the lack of consistency in the reporting. This made it extremely difficult for users to find correlations in the data in order to determine if patterns existed and, if so, whether there were also exceptions to the patterns. This cumbersome process also did not lend itself to economies of scale.
Today, companies are looking to modern BI tools that encompass both operational and traditional BI and can graphically display data and consistently across systems to provide a unified view of the essential data.
To enable these new, on-demand data aggregation requirements, operational BI is now being paired with a query-based integration technology called enterprise integration information (EII). EII provides an abstraction layer between information applications, such as BI and reporting tools, and the underlying data sources. EII abstracts disparate and distributed data sources, making them appear as though they are uniform and located in one place, delivering the results to information dashboards, customer-view portals or other information applications.
From a business perspective, EII integrates and presents information as a seamless whole. Already accessible via separate applications, portals and front-end tools, these views could encompass inventory or accounts receivables. They could also encompass booked sales data from a packaged application such as SAP, a homegrown management system, Excel spreadsheet or other data repositories, and combine them into a single report or dashboard.
With EII, operational BI extends valuable up-to-date information to those employees in an enterprise who run the actual operations - plant managers or shop floor supervisors in a manufacturing operation, or customer service or salespeople. Operational BI can offer much-needed interactive reports, enabling users to find information on the fly, and allowing them to make better decisions faster because the information can be tailored to their role and combined with historic information to provide context. Users can drill down to discover further context of the scenario. For example, an operational report for service managers could include a graph and a list chart outlining the number of service requests by day, by month and by year. The manager could then drill down to identify an individual service request for a specific customer.
Assess the Need for Operational BI
Operational BI is highly flexible and can be used throughout various areas of an organization to improve day-to-day operations. Because needs vary from company to company, here is a list of questions to help assess the need for operational BI.
You are a candidate for operational BI if your departments need to do one or more of the following:
- Combine current information with historical data to enable managers to make better decisions;
- Check the status of operations throughout the day and make appropriate adjustments;
- Handle customer and vendor requests for status information throughout the day so that the customer can improve its own operations;
- Improve performance of employees by enabling them to easily modify certain reports to show the information they need;
- Enable employees to quickly retrieve information from multiple systems to get the intra-day information they need.
For example, a sales operations manager needs to see closed orders at the end of each day from each of his company's three shipping facilities. Each facility uses a different version of SAP. The problem is that the different versions require dissimilar mechanics for accessing and interpreting the data, making it difficult and time-consuming to retrieve the information required. Or a logistics manager needs to know the order status of a particular business item, and this requires on-demand access to current data. With operational BI, the managers can now receive graphical reports that include aggregated data found in a data warehouse and is combined with current data found in their transactional system.
Examine Operational Details
Once a determination has been made that a particular business process could benefit from operational BI, the next step is to take a closer look at the details of the operation. Are homegrown systems or packaged applications being used? Where is the critical data stored, and how is it currently accessed? What reporting and analytical tools are being used - Excel, BI tools or portals? Consider how you might combine the data if you had the flexibility to provide the greatest operational insight.
A service department manager at a computer keyboard manufacturer, for example, needs to understand the status of the keyboards currently being repaired. The manager may use a system to generate a service request summary that shows all of the service requests currently in house. In addition to seeing all service requests currently in process, the manager would also like a report that shows an aging analysis by priority and by severity, so that the manager could quickly determine which requests needed to be expedited. This type of report would need to track the actual services requested, with the dates received, with the priority field (indicating importance of customer). But the existing reports are rigid, making creation of this unique report difficult.
Organizational Requirements for Project Success
In any endeavor in which changes are made to how managers and users conduct business, obtaining their buy-in is essential to ensure the full cooperation required for ultimate success.
Here are steps that can be taken to obtain buy-in:
- Establish clear lines of communication with the business sponsor and illustrate the financial benefits of the project. For example, identify who stands to gain financially if the service manager is able to clear outstanding service requests 25 percent faster than before.
- Describe the key features and benefits to the business users and communicate clearly how the project will make their lives easier. For example, order management team members should know that the operational BI system will provide them with reports highlighting current inventory levels in an easier-to-use front end, and that this will dramatically improve their responsiveness to sales people and customers. Rather than struggle with canned reports that provide limited information, they'll be able to easily modify the new reports to get the precise type of data they need.
- Map out a project schedule based on input from the business sponsor and users and involve them at key milestones:
- Analysis. Learn which tools, systems and applications users currently work with. Find out how data is presented to them. Determine which data, if delivered quicker and in a better format, would improve their productivity. Determine user needs - do they require self-service, and do they need to access current and historic information? Will data need to be distributed to users via thin clients? Also, who needs to author reports, and what are their skill sets? Finally, what overall information needs to be seen and in what type of view - graphical or just lists?
- Design. Solicit feedback on new proposed application views. How might they prefer a screen shot or report to look?
- Implementation. Keep them abreast of timing and any impact the implementation will have on their daily operations.
- Acceptance. Present the new system to them and make changes in response to user feedback.
Technological Templates Accelerate Project Implementation
The use of technological templates, or prebuilt operational reports, can dramatically accelerate project definition, implementation and acceptance. Such templates are most effective when they are based on industry best practices and have predefined metrics, columns and summaries. Reports that also provide connectivity to operational systems out of the box and that are easily modifiable to meet individualized user needs are particularly valuable for the project's success.
Figure 1 shows what one of these templates might look like. Called the Service Request Aging Analysis by Severity report (from the Cognos 8 Report Pack for Siebel), this layout enables the service manager to see at a glance the requests currently in the pipeline by their priority and severity status. For instance, if this was a cable company, the service requests could range from basic questions about new channel upgrades (lower priority) to actual problems with service quality (critical). The report shows the level of severity for each priority. Of prime concern would be those requests that are critical. According to this report, a total of six are considered critical, yet one request is one to two weeks old, and one is three to four weeks old. The manager could then choose to drill down and see the status of those that are more than one week old. Because managers are often compensated for the speed with which they can eliminate all outstanding service requests, this report would provide valuable and timely insight into the status of operations.
Figure 1: Example of a Technological Report Template
This type of prebuilt report template can be extremely valuable in each stage of the project definition phase, as follows:
Analysis. In this early stage, the report template shows users what they can actually expect. First, it reveals the type of data that can be accessed, even pulling data from multiple, disparate systems that they might not have been able to access previously. The report also gives an idea of the different ways that the data might be combined, helping users to imagine new ways of mixing data that could greatly help them in day-to-day operations. Finally, it allows them to imagine ways in which the data could be presented for the greatest impact.
Design and Implementation. For the IT managers who will be building the operational BI system, the templates provide a tremendous starting point as 80 percent of the solution is already built and includes self-service features for end users. For example, a good template should sit on top of a server such as the Composite Information Server that leverages all of the modeling, query building, optimization, publishing (SQL and Web service), caching and management capabilities. It also should utilize adapters - or connectors - that are ERP-specific and have been written against published interfaces sanctioned by the packaged application. For instance, it would include the BAPI calls and RFC function invocations that are standard for data retrieval within SAP. If interfacing with Siebel, it would have adaptors that adhere to Siebel's object managers. Implementing operational BI will provide some companies with an opportunity to migrate away from long-standing, yet less-than-ideal methods of delivering reports, such as coding or using a report writer. A feature-rich BI application enables self-service, provides easy-to-use authoring tools, and improves data access capabilities - simplifying deployment of operational BI and other reporting needs.
Acceptance. Very few surprises should occur at this stage, as the completed design should map closely to what the users saw in the beginning of the project. Any changes are typically minor and can be easily resolved through modifications to the design.
Assessment of the Operational BI Project after Implementation
Once the new system has been operational for three to six months, users will most likely have ideas and suggestions on how the system could better meet their needs. Suddenly, capabilities they had never before thought possible will now become achievable. If the system was designed and implemented properly, users should be seeing results very much in line with the projections first established in the analysis phase - and these should be quantified and presented to the business sponsor. This is also a good time to make improvements and to dynamically change the reports as business conditions change.
Empowering Those on the Front Lines
With technological templates, companies now have a unique opportunity to rapidly assess, design and implement powerful operational BI capabilities utilizing the same BI tools they have invested in for their operational data store or data warehouse. Managers who operate on the front lines - interfacing with customers, checking inventory, managing shipping dates or tracking financials on demand - will, for the first time, have access to critical data all in the same report that was once located in multiple, difficult-to-reach, disparate sources. What's more, the data will be fresh, culled daily and presented in dynamic information dashboards in visually exciting ways. As the project progresses, the reports can easily be modified without coding to meet the rapidly changing needs of the end users. Ultimately, this will enable managers to resolve outstanding issues faster, serve customers more quickly, and increase the speed and accuracy with which they serve the business.
For more information on related topics visit the following related portals...
Business Intelligence (BI),
Project Management/Tool Selection and
Query & Reporting.
David Griffin is director of Market Development at Cognos. Griffin looks at changes in the performance management market and works with product management to evolve products and solutions. He led the product conception and management of Cognos Report Packs, the industry's leading out-of-the-box solution for ERP reporting on the Cognos 8 BI platform. He has worked at Cognos for seven years following an acquisition. He may be reached at firstname.lastname@example.org.
Jason Monberg is vice president of Application Products at Composite Software. Jason has over 12 years of experience developing and managing enterprise software applications. He has been focused primarily on data management and integration for the last seven years. Currently, he is responsible for the Composite Application Views product line that allows users to query SAP, Siebel, PeopleSoft, Oracle E-Business Suite and Salesforce.com. Previously, Jason founded and raised capital for three software startups. He also has experience in software engineering and architecture, product management and marketing and consulting. He may be reached at email@example.com.