Chapter 4
Exploration and Reporting
The world of data exploration is ruled by reporting and visualization . . . thoughts on slice and dice
It was 2006, and I was in India. In a daze of humidity and heat, I was wondering, “How can we implement business analytics functionality in a financial enterprise in a country like India?” As I passed underneath a tangled web of electrical wires, I heard my name. “Hello, Al!” A man was following my winding path through a crowd of brightly colored saris, moving effortlessly behind me. I turned around and walked to greet him. I had spent time with him over the phone in the past few weeks.
At that moment, I thought to myself, “I’ve built a brand-new data warehouse with lots of data stored neatly inside. What am I going to do with it?” Climbing through the beautiful and ancient city of Delhi, big data are on my mind.
I am here in Delhi to observe a colleague of mine, Sanjiv, teach his staff about business intelligence, a.k.a. reporting. He will be leading a class on big data, emphasizing that the first and most important use of new analytical platforms is effective visualization. Alternatively, reporting can be used to support statistical analysis, that is, modeling.
I shook Sanjiv’s hand and after an exchang of pleasantries, we launched into the topic at hand. “Current reporting is definitely not our father’s reporting. I remember waiting in my office at Blue Cross and Blue Shield Association in Washington, D.C., for a huge stack of paper delivered on a hand truck in the 1980s. Today, analysts use in-memory data that delivers interactive graphs on Web portals where business analysts can slice and dice data enriched with images and graphs. They reach out to the enterprise via SMS, videoconferencing, and/or e-mail. They get their information immediately on their mobile devices like iPads and iPhones. What a difference!”
Sanjiv smiled knowingly. “Yes. The Internet and mobility have revolutionized the way successful businesses communicate with their customers, vendors, and suppliers. New Web-based platforms help decision makers gain insight from their data, discover what customers want and need, and enhance product offerings for long-term profitability. More important, the distribution of information is much cheaper.”
I laughed. “Every day, data managers face the challenge of managing terabytes or petabytes of corporate data. Big data projects are blooming everywhere. Information flows into corporations from customers, vendors, and suppliers. Managing, accessing, and understanding these data are complex processes. We are drowning in data!”
Sanjiv wiped his brow. “I work hard to teach my staff that competitive companies are selecting innovative solution to leverage their limited resources and to deliver fresh, relevant information and tools for understanding big data as quickly and easily as possible. Cloud technology is helping deliver high-quality data to key decision makers to perform whenever and wherever they need them. This evolving technology is providing decision makers with the information and knowledge they need to understand their customers’ experience, predict customer needs and behavior, and ultimately grow their businesses.”
I added, “Definitely! Corporate acceptance of Web-based analytics is also growing. Decision makers are evaluating competitive technologies, corporate objectives, and resources and are deploying analytical solutions for applications from customer relationship management [CRM] to data mining. Cloud CRM analytics have experienced exponential growth in the last few years.”
Sanjiv and I get to the hotel lobby. We sigh with relief as we enter the air-conditioned calm and head upstairs to try to find the meeting room. A colleague of Sanjiv’s, Samir, is speaking at the front of the large conference room.
VISUALIZATION
Samir reads the following few lines from a piece of paper he has in his hand: “Understanding Web-Based Analytics.” He takes a sip of water. “Visual analytics can deliver standardized analytic dashboards to end users’ desktops via simple Web browsers or similar technology. Using a Web browser as a ‘viewer’ allows users to access analytical software and create interactive graphics to discover important trends in their entire data loaded in memory. In such analytic system, users access data from a central repository. Users analyze data specific to their needs and then create attractive graphics that will translate the results to associates, from senior executives to field managers.”
Figure 4.1 shows a typical SAS® Visual Analytics dashboard.
Figure 4.1 SAS® Visual Analytics Dashboard
Source: Copyright © 2014 SAS Institute Inc. SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc., Cary, NC, USA.
A hand shoots up. “For example, can marketers analyze customer feedback generated from a corporate hotline number? If so, these data could then be analyzed, graphically displayed, and delivered to product development, manufacturing, and senior executives who can make adjustments in real time as appropriate.”
Samir responds, “Good point. To answer your question, yes they can. Let me ask another question: what are some of the advantages of visualization then?”
A student replied, “A lot of people are visual. Don’t forget that a picture is worth a thousand words. Visualization offers decision makers several advantages including real-time access to data and interactive slice and dice as well as drill-down graphic capabilities.”
I also raised my hand. “Furthermore, new visualization and reporting technologies help companies implement best practices and facilitate information sharing and collaboration among decision makers at low price points without compromising scalability. New reporting technology advances promise greater interactivity and flexibility, facilitating the inevitable transition from data acquisition to in-depth analysis.”
Samir lit up. “Oh hello, Al. I didn’t see you back there. Yes, today data managers need to help cleanse and manage an enormous amount of data to feed reporting and visualization systems that deliver an initial set of information to corporate decision makers.”
Everyone looked back at us, and Sanjiv continued, “Yes. In many instances, a cloud computing platform offers a good opportunity for systems integration and secured multilevel access to critical information.”
Samir introduces us. “These are two big data masters, Al and Sanjiv. They taught me everything I know.”
I waved. “Don’t forget that users can access and visualize several databases from a centralized location or analyze sections of data, if required.”
I scanned the crowd of nodding heads. Sanjiv continued, “Believe it or not, previously, analytical tools and research databases were delivered to users on a customized CD-ROM.”
Sanjiv chuckled. “The new visualization and reporting technology provides decision makers with real-time data sometimes stored in memory—it has become increasingly cheap—and associated with powerful analytical tools to deliver fresh, relevant results. These tools allow the detection of patterns immediately.”
Sanjiv went on, “Yes. Another significant advantage is providing a mechanism for secure delivery of data. In the past, data managers were reluctant to provide users with sensitive data in a CD-ROM format because of obvious security risks. Now data managers can protect sensitive data while allowing key decision makers access to sections of the database from a password-protected, centralized server using security protocols. Other advantages include version control and efficient processing of changes to decision makers’ analytical and data acquisition tools. This contributes to data consistency across the organization and also fosters collaboration.”
Samir brought the class to an end, and I suggested to Sanjiv and Samir that we all go up on the rooftop.
I could not help talking business with two of India’s brightest minds. “With Web-based applications, users can access current data to make decisions at e-speed. Increased velocity in the decision-making process enables decisions to be made more quickly and corporations to become more responsive to consumer needs. That helps organizations cope with the velocity, variety, and volume of new data arriving.”
Sanjiv continued my thought. “Faster, more accurate decision making can help our companies meet emerging customer needs or avoid costly mistakes. For example, real-time data fed from manufacturing plants can help plant floor managers make production adjustments ensuring that high-quality products are delivered to consumers. This is invaluable here in India.”
Samir added, “Our senior executives can gain immediate insight regarding emerging trends or anticipate potential problems with their products and services, thus becoming more responsive to changing market conditions.”
I agreed. “New, interactive technologies, such as visual analytics using in-memory capabilities, are providing decision makers more information at lower costs.”
Samir smiled and asked, “How is visualization technology moving graphic presentation from a static appearance to an interactive format?”
I answered slowly, “One very common way is by using Java applications, delivered via a Web browser. These applications can be programmed to interoperate with other applications and allow the user to drill down into the data or to create hyperlinked graphics to relevant documents, giving decision makers more information about each data point. This discovery process becomes faster. In reality, the visual analytics systems are question generation systems rather than answering systems. Of course, some of the questions become actionable items immediately. Otherwise it would be pure academics.”
I went on, “People collaborate because background information or comments can be attached to each data point in a chart. For example, an analyst can view news releases attached to points within a times-series chart or more detailed regional information could be attached to data points in a map. Also, you can open a visual analytics dashboard with your peers on a Webex session and get a lot of work done.”
I drew on a piece of paper a model of actionable data discovery (see Figure 4.2). “This model represents the data integration needed to assemble data. It is sort of a data factory. From the factory, we extract data for a single purpose: to answer a research question. A researcher, domain expert, may take a look at the selected data and discover some interesting trends or events. At that moment, there are two main possible tracks: One, some discoveries will merit immediate action or two, some discoveries will pose new questions. New questions may need additional data, and the cycle goes on and on. That’s what this graph shows, the iterative nature of data-based discovery.
Figure 4.2 Model of Actionable Data Discovery
Sanjiv asked, “So, could new visualization tools that are rendering graphics and relevant information on the fly allow analysts and decision makers to visualize real-time data changes instantly?”
“Of course, if that’s what is needed,” I replied. “System integrators and data managers are constantly challenged to build solutions that are timely, flexible, and scalable to meet the needs of growing organizations. Delivering graphs over the Web is the perfect solution. A picture is worth a thousand words.”
Samir added, “The days of fitting individual PCs with the latest software are gone. It is much cheaper and more efficient to deliver technology based on a simple Web browser, which makes it easy for IT departments to install and maintain the application and for decision makers to access the tools they need to gain insight from their data. In addition, as new employees are hired, the system has the flexibility to add new users easily and training issues are minimal.”
CLOUD REPORTING
I was late for my morning class. When I arrived, Sanjiv was saying “Cloud reporting offers an opportunity to customize visualization by department or region. As a company grows, reporting can be customized to meet new organizational or directional needs. For example, if a packaged-goods company adds new plants in new countries, the reporting system can be updated within minutes.”
A student asked, “As companies expand or diversify, can additional servers be added to meet new performance and scalability demands?”
Sanjiv replied, “Yes, and cloud delivery creates a familiar environment, making it easy to train new employees. Delivering reports via the cloud saves time and money for IT managers who don’t need to update individual software packages on each employee’s computer.”
Another student asked, “How does the distribution of reports from a central server allow business intelligence administrators to deliver a single truth by department, geographic location, or need?”
Sanjiv smiled, “This deployment method makes standardization easy for companies interested in maintaining best practices. The reporting software could ensure that high-quality products are developed and varying standards are enforced. Administrators can provide and receive feedback in the system from decision makers and make changes to the visualization software as necessary. Since changes can be deployed enterprise-wide within minutes, the BI administrator can be sure that everyone is using the same application.”
The student murmured, “Wow. Cloud reporting is fast and economical, and it can help decision makers explore business results seamlessly across borders.”
“Yes. For example, retail bankers can collect data and analyze results globally using a prescribed set of standards. Standardization leads to improved communication and comparison of results. Modern visualization tools help communicate these standard results to senior executives and research team members globally. Advanced visualization technology provides the ability to drill down for additional information immediately. OK, let’s break for lunch!”
Sanjiv kept talking after class was done. “Al, with cloud reporting, decision makers have access to greater processing power because their scalable environment with several tools can perform analysis on the server. Greater server capacity offers decision makers greater access to advanced reporting and basic predictive analytics, which can then be distributed using the Web.”
We jumped into a cab outside. Sanjiv was still talking. “As needs change, administrators can add or remove software tools from the system to provide additional functionalities and increase data storage. Greater processing and access to more powerful analytics can lead to a more in-depth understanding of the data, creating greater predictive value and better decision making. Collaboration and information sharing are skyrocketing. Attractive, well-designed, multidimensional graphics and dashboards combined with interactive technology will improve insight and information sharing.” He continued: “Sharing results from the visualization and reporting systems to personal productivity desktop packages is not difficult anymore, and sharing results within the enterprise or with international associates is simpler using Internet-based applications.”
Sanjiv started to slow down. “Cloud reporting and visualization software components are providing decision makers with greater access to data to gain insight into their customers’ buying patterns and other critical customer insight.”