Business intelligence (BI) software solutions have been evolving steadily to become more usable by business analysts. Given the amount of information that is collected and stored electronically, the incentive is strong to get value from it. But to do so, business users must be able to access it themselves, not wait for a report from IT.
The city of Charlotte, N.C., known for its banking industry, is consistently rated as one of the best places to live and work in the United States. The city government's key business units support a multitude of services that are provided to Charlotte's half-million residents. To manage and improve those services, the city must monitor the performance of its business units across a wide range of activities.
BI tools have proven to be an effective way to analyze performance and provide dashboards, but the development time using some of the available BI tools was unacceptably long. That led the Charlotte's Business Support Services/Information Technology (BSS/IT) leaders to search for a software product that could present visualizations quickly and help users make valid, data-driven decisions.
Reaping return on investment
The software would need to empower business users, be easy to use and have a strong set of built-in visualization tools. Those capabilities, plus an affordable price, landed Tableau Desktop from Tableau Software at the top of BSS/IT's list. "We now have approximately 40 users licensed for Tableau Desktop," says Jim Raper, manager of the data administration team for the city of Charlotte. "We have also installed a similar number of the user version of Tableau, which is Web-based, along with hundreds of copies of the free reader."
One of the attractive features of Tableau is the ability to easily bring data in for analysis. For instance, users can drag and drop fields into Tableau for analysis, and the software easily connects to other data programs. "Connecting to Oracle, SQL Server, Access and Excel is routine," says Raper. "We just have to hit the ‘Connect to Data' button and provide the name of the database and the table we want."
An example of the type of analysis that provides a compelling case for Tableau's ROI was a comparison of the impact of oil additives on the fuel mileage of city refuse collection trucks. Raper explains, "We put the data into Tableau, and in 15 minutes we discovered that one of the additives was not improving mileage. This piece of data helped save the city $50,000 in future expenditures."
In the past, business users might look at a report generated by BSS/IT and realize they had not asked quite the right question. A second iteration would follow, with the user waiting in the queue for the updated analysis report. With Tableau, users who want to tweak their initial analysis are able to do so themselves. "Putting a top-quality tool in the hands of non-IT users helps them be more productive," Raper says, "and my staff is now able to focus on the more challenging questions that might be problematic for business users."
The visualization capability of Tableau is another strong point, according to Raper. "The default visualizations are generally best practice from the viewpoint of what the eye can assimilate," he says. Dashboards are also quick and easy to create. Data can be shown on a map and color-coded to indicate performance levels.
Live, Web-based training is offered by Tableau twice a week at no cost. "Having all the training available for free is quite unusual," says Raper. "Since training budgets are often the first thing to be cut when budgets are tightened, that has been very helpful." Occasional users can always go back for a refresher. "Between the online training and a half-hour session of side-by-side mentoring, our non-technical users are ready to go," he adds.
To help polish users' skills, Charlotte's BI community stages a BI Olympiad every two years, in which business units across the city compete. "Each team is given the task of picking a real business problem, finding the data and presenting it in a dashboard along with a write-up for a decision maker to evaluate," explains Raper. In the final round, teams are presented with a critical scenario the city needs to contend with: most recently, the simultaneous arrival of a hurricane, gas crisis and bioterrorism. The teams have 24 hours to come up with a BI solution that provides information useful in responding to the situation, presented in a visually effective manner. "Although the teams can use any BI tool they want," says Raper, "the top three all used Tableau, because the analyses could be done so quickly."
Designed for broad use
Tableau Software emerged from software developed at Stanford University for a defense project. Its chief scientist and one of Tableau's co-founders, Pat Hanrahan, was a founding employee at Pixar, where he was chief architect of an innovative rendering protocol for graphics. The combination of query capabilities and strong computer graphics for visualization have helped Tableau take root in about 6,000 companies in nearly 100 countries.
"Because it's so easy to use, people enjoy working with Tableau," says Elissa Fink, VP of marketing at Tableau. "Our customers range from Fortune 100 companies to mom-and-pop shops. Companies do not need a dedicated staff to get value from the product; on the contrary, it is designed for broad use." A customer of Tableau's before going to work there, Fink had been stymied in her efforts to extract meaningful data from traditional BI products. "With Tableau, users can get started right away, and gain insights from the very beginning," she says.
Automating the scorecard
Companies that move from manual processes to a BI system experience a quantum leap in their ability to provide timely analyses of operational data. ResCare provides services to more than a million people with special needs each year. To monitor its performance, the company prepared a scorecard that covered major business areas, including financial data, risk management and customer satisfaction. One employee gathered information from numerous sources in different formats and media, including both paper and electronically stored information, and aggregated it to produce the scorecard. The process took up to eight weeks and was completed on a quarterly basis.
Wanting to automate the process, ResCare hired a consulting firm to provide an intensive, objective assessment of available BI options. The study took about six months and narrowed the choice to three vendors. After carefully considering a large set of decision factors, ResCare selected two products from Oracle, Oracle Business Intelligence Foundation and Oracle Business Intelligence Application.
The implementation was done with two primary goals in mind. "We wanted to have a much quicker turnaround time for our analyses," says Joe Lichtefeld, VP of application services at ResCare. "Seeing data two or three months later did not provide us with the decision-making capability we needed. At the same time, we wanted the analyses to reach more people so they could benefit from the information."
Financial data was the first area addressed by the system. "As a longtime Oracle customer, we were able to leverage a lot of data from our enterprise system," Lichtefeld says. "Some of the data is still coming from spreadsheets, but rather than being dealt with manually, the spreadsheets are put on a shared network drive and the BI software picks it up from there."
The Oracle BI solution displays data on a dashboard that is tailored to the employee's role. "The biggest benefit that we are seeing, which is what we had hoped for, is that executive directors of the individual units can see data at the level of their operations, the regional manager can view it from that level and so on," adds Lichtefeld. In the past, the information was available in theory, yet not readily accessible. "Now, in 20 minutes, our employees can look at the main metrics for their business activities," he says.
A dashboard shows the five business areas, with a green indicator if all performance targets are being met, and red if there is a problem. If trouble is indicated, the user can drill down to find the specific source. The dashboard layout is identical for each user, but the default data shown is appropriate to his or her level. Some of the performance measures are shown as bar charts and others as pie charts.
Initially, the plan was to focus on the financial data portion of the scorecard, because the BI system could connect to the Oracle data quickly. But before the system was launched, ResCare decided that it wanted to include all the elements of the scorecard. "We are a service business," Lichtefeld explains. "We believe that providing quality services is equally important with cost-effectiveness, so we delayed the project in order to incorporate all the metrics, including customer satisfaction."
The data is refreshed from the data warehouse each evening, to provide daily updates for one of the lines of business. For other units, the updates take place at varying times. "Rather than seeking real-time monitoring, we aim at ‘right time' measures and a single version of the truth," Lichtefeld says.
Oracle's Business Intelligence Suite offers the advantage of being able to integrate readily with financial data contained in other Oracle systems. Its scalability and strength in handling complex analytics also make it attractive to organizations that want the options of extending their BI applications over time.
Real-time Web data for enhanced BI
Increasingly, information that could be valuable to have in BI systems is available on the Web, but extracting it for inclusion in data warehouses is difficult. Kapow Katalyst Platform from Kapow Software is an application integration product that performs extraction, transformation and integration from any Web-based data source without needing an API. "We work with BI vendors to get data that is important but cannot be obtained through traditional means," says Stefan Andreasen, CTO and founder of Kapow Software.
Kapow Katalyst provides real-time information for applications such as competitive and financial intelligence. "Katalyst can automate any repetitive tasks on the Web, such as navigating through pages on a website and extracting and transforming pricing data for cameras, for example," Andreasen says. "A company can then use that information to adjust its own prices to be more competitive." Kapow Katalyst can also seek out and extract qualitative data from social media, imposing metadata such as date, title and author name as well as removing "noise" such as ads and navigation.
As another example, Fiserv, which provides technology solutions to financial institutions, is using Kapow Katalyst to automate access to external financial accounts. Prior to using Katalyst, four people were logging into 300 Web banking
applications and manually copying data into spreadsheets. The process took a week and ran the risk of errors. By using the solution, each bank was in effect converted into an ETL feed and the data was available in near real time.