Learning, Exploration, and Discovery with Visualization

Data visualization enables learning that is the basis for continuous improvement. When companies, political parties, sports teams, or fund-raising agencies invest in marketing programs, campaigns, promotions, special events, or other projects, they use visualization to learn something from them. Visualization is also used as a data explorer and data discovery tool. Companies, such as Safeway and PepsiCo, are discovering new relationships and learning how to improve performance using data visualization in all types of industries and governmental agencies. Enterprise visualization apps for Androids, Apple iPads, and Surface tablets are replacing static business reports with real-time data, analytics, and interactive reporting tools.

Examples of Visuals

Examples of visualizations include dials, charts, graphs, timelines, geospatial maps, and heat maps. The tricolor heat map in Figure 11.4 instantly alerts the viewer to critical areas most in need of attention. Visual displays make it easier for individuals to understand data and identify patterns that offer answers to business questions such as “Which product lines have the highest and lowest profit margins in each region?” Interactivity and drill-down capabilities are standard features that make visualization even more valuable. Two other types of heat maps, created in Tableau Desktop, are shown in Figure 11.5; both heat maps are based on the same data set. Notice that the way in which the data are visually displayed depends on what you want to learn or convey.

Human expertise is an essential component of data visualization (see Figure 11.6). A common mistake organizations make is to invest in the analytics foundation—tools, quality data, data integration, touch screens—but overlook the most crucial component, which is the users’ ability to interpret the visual reports and analyze them correctly.

Data Discovery Market Separates from the BI Market

According to Gartner Research, the data analytics market has split into two segments: the traditional BI market and the newer data discovery market. Data discovery software had previously been viewed as a supplement to traditional BI platforms. Now it is a stand-alone alternative to BI. This split occurred because today’s data discovery technologies provide greater data exploration and ease of use to help users find answers to “why” and “what if” questions through self-service analytic apps. The split is another example of pushing analytics onto the computers of business workers. IT at Work 11.1 describes the trend at IBM.

Data Discovery Offers Speed and Flexibility

Data discovery is expected to take on a greater role in corporate decision making. Companies are investing in the latest data discovery solutions largely because of their speed and flexibility. Experts and novices can collect data quickly from disparate sources and then explore the data set with easy-to-use interactive visualizations and search interfaces (Figure 11.7). Drill-down paths are not predefined, which gives users more flexibility in how they view detailed data.

A powerful feature of data discovery systems is their ability to integrate data from multiple data stores and identify data types and roles. See Tech Note 11.1. While data are being loaded into the program, the software automatically extracts and organizes them by data type. Software may also extract and organize terms from unstructured content, such as texts, e-mail, and PDFs, and create tag clouds. Figure 11.8 shows an example of a Word cloud that give users a quick way to evaluate the most aspects of SCM and start to make discoveries.

How Is Data Visualization Used in Business?

The ultimate goal of data analytics is to drive profits, and often that depends on learning how to manage assets, such as inventory, or engage customers in a smarter way. Collecting data is relatively easy. Making sense of that data is not. Here are examples of how companies and/or entire industries are using data visualization and interactivity to improve decision speed and performance often with mobile displays.

The latest data visualization software addresses issues associated with processing big data by speeding up data discovery and returns the visualization within an appropriate timeframe, in an easy-to-understand format. BI and data visualization vendors are working to assist business analysts and nontechnical users in determining how best to display these massive amounts of data.

Data Visualization Tools

A number of vendors offer data visualization software. The following list describes a few of these. Many of these vendors offer a free trial of their software on the websites.

• SAS Visual Catalyst has an intelligent autocharting feature that automatically presents the most appropriate visualization of a specified data set based on the amount and type of data being analyzed. By building hierarchies “on the fly,” interactively exploring data, and displaying data in different ways to answer specific questions or solve new problems, these new data visualization products relieve the user of constantly having to rely on assistance from the IT department when they want to change the ways in which the data is displayed. For example, if you have a billion rows of data, it would be impossible to see so many data points on a scatter plot. However, a box plot would convey the information that you need.
• Birst combines capability, scale, and data governance that IT needs with the agility, speed, and usability of consumer-grade desktop tools. Birst’s adaptive user experience offers users a wide range of self-service, data analysis, and presentation options, and its cloud architecture allows users to instantly share findings and data across its supply chain.
• QlikView distinguishes itself from other BI tools through its unique inference engine that automatically maintains data associations. The inference engine works much the same as the human brain, in that when it “thinks” of a data set, it is reminded of all things related to it. Qlikview also offers a simple, Google-like search that works in an associative manner, producing results for the phrase and also for things commonly related to the phrase. For example, if a user wants contact information for a salesperson in an organization, and only knows the product that the salesperson specializes in but didn’t know the person’s name or the company, Qlikview could produce the desired results.
• IBM SPSS Analytic Catalyst has made sophisticated analytics accessible. Analytic Catalyst enables business users to conduct the kind of advanced analysis that had been designed for experts in statistical software. The software fast tracks analytics by identifying key drivers, selecting an appropriate model, testing it, and then explaining the results in plain English. See the YouTube video titled “IBM SPSS Analytic Catalyst” for an overview. The tool condenses the analytic process into three steps: data upload, selection of the target variable (the dependent variable or outcome variable), and data exploration. Once the data are uploaded, the system selects target variables and automatically correlates and associates the data. Based on characteristics of the data, Analytic Catalyst chooses the appropriate method and returns summary data rather than statistical data. On the initial screen, it communicates the so-called top insights in plain text and presents visuals, such as a decision tree in a churn analysis. Once the user has absorbed the top-level information, he or she can drill down into top key drivers. This enables users to see interactivity between attributes.
• IBM Watson Analytics is a cloud-based data visualization tool that guides data exploration, automates predictive analytics, and enables effortless creation of dashboards and infographics.
• Tableau is one of the easier data discovery tools to implement, requiring just basic database information to connect it to the target data sources. With a new in-memory database engine, such tools are developing the power to perform big data analytics. Despite data visualization advancements, data integration between data sources can still be very challenging.
• Roambi Analytics is a leading mobile reporting and data visualization app designed for iPads and iPhones. The app can take data from most sources, including Box, Google Docs, spreadsheets, BI systems, databases, and Salesforce.com, and transform them into interactive data visualizations. Roambi has a worldwide customer base of Global 500 companies and small and medium businesses across industries, including telecommunications, biotechnology, pharmaceuticals, consumer technology, and packaged goods.

Concept Check 11.1

Mashup Architecture

Technically, a data mashup is a technique for building applications that combine data from multiple sources to create an integrated experience. As techniques for creating mashups became easier, companies started using them to build enterprise mashups that supported their business models. Tech-savvy managers realize that they can use mashup apps with their existing data and external services to provide new and interesting views on the data.

Figure 11.9 shows the general architecture of an enterprise mashup app. Data from operational data stores, business systems, external data (economic data, suppliers; information, competitors’ activities), and real-time news feeds are integrated to generate an enterprise mashup.

Why Do Business Users Need Data Mashup Technology?

Business users have a hard enough time identifying their current data needs. It is not realistic to expect them also to consider all the new sources of data that might be made available to them and the analyses they might do if they had access to that data. With traditional BI and data warehousing systems, data sources have to be identified, and some understanding of data requirements and data models is needed.

Realizing that there will always be data needs that cannot be anticipated, the question is whether IT should be in the middle of supporting those requests? Providing business users with self-service enables them to meet their needs more quickly. They also have the opportunity to explore and experiment.

Enterprise mashups improve operational efficiency, optimize the sales pipeline, enhance customer satisfaction, and drive profitability. Within government, mashups have positively impacted strategic areas such as citizen engagement and satisfaction, financial transparency, project oversight, regulatory compliance, and legislated reporting. A summary of enterprise mashup benefits is given in Table 11.2.

Enterprise Mashup Technology

Mashup technology leverages investments in both BI tools and interactive technologies. BI systems are very good at filtering and aggregating huge data volumes into information. With mashup technology, for example, users can filter down the data based on their needs so that only the information needed is provided by the available data services. Tech Note 11.3 describes mashup self-service.

Concept Check 11.2

Dashboards are Real Time

Dashboards are often mistakenly thought of as reports consisting of various gauges, charts, and dials, but the purpose of business dashboards is much more specific and directed. The purpose of dashboards is to give users a clear view of the current state of KPIs, real-time alerts, and other metrics about operations. Dashboard design is a critical factor because business users need to be able to understand the significance of the dashboard information at a glance and have the capability to drill down to one or more levels of detail. Having real time, or near real time, data is essential to keep users aware of any meaningful changes in the metrics as they occur and to provide information for making decisions in real time. Users can take corrective actions promptly.

It’s easy to see in Figure 11.11 how color-coded displays can quickly inform the user of the status of KPIs.

How Operational and Strategic Dashboards Work

Dashboards are custom programmed to automatically and securely pull, analyze, and display data from enterprise systems, cloud apps, data feeds, and external sources. They work by connecting to business systems, such as accounting software, ERP, CRM, SCM, e-mail systems, website analytics programs, and project management software via APIs. IT at Work 11.2 describes dashboards in action at Hartford Hospital. Tech Note 11.4 lists vendors that offer free trials of dashboard software.

Benefits of Digital Dashboards

The interrelated benefits of business dashboards are as follows:

1. Visibility Blind spots are minimized or eliminated. Threats and opportunities are detected as soon as possible.
2. Continuous improvement A famous warning from Peter Drucker was “if you can’t measure it, you can’t improve it.” Executive dashboards are custom designed to display the user’s critical metrics and measures.
3. Single sign-on Managers can spend a lot of time logging into various business systems and running reports. Single-sign-on dashboards save time and effort.
4. Deviations from what was budgeted or planned Any metrics, such as those listed in Table 11.3, can be programmed to display deviations from targets, such as comparisons of actual and planned or budgeted.
5. Accountability When employees know that their performance is tracked in near real time and can see their results, they tend to be motivated to improve their performance.

Concept Check 11.3

Geocoding

In many cases, locations are already in existing data stores, but not in a format suitable for analytics. A simple process called geocoding can convert postal addresses to geospatial data that can then be measured and analyzed. By tapping into this resource, decision-makers can use the geographic or spatial context to detect and respond to opportunities.

GIS Is Not Your Grandfather’s Map

Unlike a traditional flat map, a GIS-generated map is made up of many layers of information that provides users different ways to view a geographic space (Figure 11.13).

Imagine for a moment that you are a regional sales manager who needs to view sales data for one of your 75 stores distributed throughout the State of South Carolina. On a flat map of South Carolina, if you looked at retail store #50, you would see the name of the store and a dot showing where it is located on the map. However, if you view a GIS map of the United States on your computer, smartphone, or tablet, you can hover over South Carolina and when you click on retail store #50, up pops the store’s location, store manager’s name and phone number, weekly and monthly revenue, product categories, a photo of the storefront, and a virtual tour. As a highly paid, busy regional sales manager, this saves you time and your company money, increasing organizational effectiveness and efficiency.

Infrastructure and Location-Aware Collection of Geospatial Data

The infrastructure needed to collect geospatial data continues to expand. Cellular and Internet service providers, sensors, Google Earth, GPS, and RFID systems know the location of each connected user or object. Foursquare, Google Maps, and other mobile apps rely on GPS locations. With the Shopkick app, Macy’s can track a shopper’s every move within one of its stores and send the shopper notifications about deals and items of interest. iBeacon is a feature available in iOS 7 devices that uses a low-power Bluetooth transmission to broadcast a user’s location. iBeacon allows Apple, or app developers leveraging Apple technology, to track users inside buildings where satellite transmissions may not reach.

Similarly to Macy’s, businesses can motivate customers to download a location-tracking app. Using GIS can help businesses target their customer markets more effectively and dynamically by engaging with them in real time.

Applying GIS in Business

GIS tools have made significant contributions to decision making in finance, accounting, marketing, and BI. Business applications include the following:

• Analysts can pinpoint the average income in areas where the highest performing stores are established.
• Retailers can learn how store sales are impacted by population or the proximity to competitors’ stores.
• A retail chain with plans to open a hundred new stores can use GIS to identify relevant demographics, proximity to highways, public transportation, and competitors’ stores to select the best location options.
• Food and consumer products companies can chart locations of complaint calls, enabling product traceability in the event of a crisis or recall.
• Sales reps might better target their customer visits by analyzing the geography of sales targets.

With current GIS, geospatial, and geocoding technologies and platforms, GISs can be easily incorporated and managed within data analytics and visualization software.

With the GIS moving into the cloud, developers of enterprise applications based on SAP, Microsoft Office, SharePoint, MicroStrategy, IBM Cognos, and Microsoft Dynamics CRM are using it to create a wide range of mobile applications.

Concept Check 11.4