Visual Reporting and Analysis: Seeing Is Knowing

概述 | 本文内容: 

Data visualization is becoming an increasingly important part of business intelligence. In a survey conducted by TDWI, 74% of respondents credit data visualization for a “very high” or “high” increase in business user insights. Business users of all levels can now answer their own questions by manipulating fast, easy-to-use visual analytics dashboards.

In this paper, TDWI shares the best visual reporting and analysis methods along with useful report writing examples, so any business user with a question—regardless of analytics training—can observe valuable insights from their data.

Data visualization technology can help you extract insights—fast. This paper will teach you why, and how you can add this useful concept to your own data discovery strategy.

We've also pulled out the first several pages of the whitepaper for you to read. Download the PDF on the right to read the rest.

Executive Summary

Data visualization is increasingly an essential element of business intelligence (BI). No longer restricted to specialized applications, data visualization in the form of charts, maps, and other graphical representations is enabling business users to better understand data and use it to achieve tactical and strategic objectives. Moreover, data visualization is prompting a cultural shift toward more analytic, data-driven business and operations by empowering users to explore, in a graphically inviting medium, data that was previously available only in tabular reports.

This TDWI Best Practices Report, which is based on a Web survey of BI professionals and interviews with BI practitioners
and experts, finds that data visualization is in the middle of a remarkable growth phase. It also reveals that data visualization contributes impressively to improvements in business user insight and productivity, as well as usage of
dashboards (the preferred medium for data visualization). For instance, 74% of our survey respondents credit data visualization for a “very high” or “high” increase in business user insights.

But data visualization is never a plug-and-play solution, and one size does not fit all. Dashboard design and usage can and should vary by types of users (for example, executives versus front-line staff), purpose (strategic, tactical, operational), and industry and organizational culture (a healthcare organization versus a clothing manufacturer). Customization, collaboration, and iteration are required for organizations to operate interactive visual reporting and analysis solutions that deliver maximum benefits.

Research Methodology

Focus. This report is designed for BI professionals and business users who develop and use business dashboards and other means of visual reporting and analysis.

Methodology. This research for this report includes a Web survey of BI professionals and business users that TDWI conducted in the summer of 2010, as well as in-depth interviews with BI practitioners and data visualization solutions providers. The survey drew responses from 222 respondents. From these, we excluded incomplete responses and respondents who identified themselves as academics or vendor employees. The completed surveys of 210 respondents form the core data sample for this report. Throughout the report, rounding and multi-choice questions account for percentages that do not total 100%.

Survey Respondent Profile. More than three-quarters (76%) of respondents are BI or IT professionals. More than 70% are based in the United States. Financial services, software/Internet, and healthcare are the most represented of more than two dozen industries. The following charts highlight key characteristics of the respondent pool.

The Evolution of Data Visualization

Long before there were servers and laptops, dashboards and drilldowns, OLAP and AJAX, heatmaps or histograms, there was data visualization.

The world’s first bar, line, and pie charts of any renown are believed to have been devised in the late 1700s by a Scottish engineer and political economist, William Playfair (1759–1823). Considered the founder of graphical data representation, Playfair produced a number of visualizations on economic and political issues of interest to the British empire. For instance, Playfair visually portrayed the growth of British debt from 1688 to 1800 in a graphic called “Chart of the National Debt of England”

Of course, Playfair labored with a crude pen and ink to produce his illustrations. He had no photocopier, scanner, or Web site to showcase and propagate his work—only a small number of books. Given the toil involved and the absence of a mass audience display media, it’s perhaps unsurprising that Playfair’s innovation had little impact, and that data visualization would not become widespread until the advent of computing.

An observation Playfair made in the 18th century is true in the 20th century as well: “On inspecting any one of these Charts attentively, a sufficiently distinct impression will be made … and the idea which does remain will be simple and complete.”

Mass Market

Although this report focuses on data visualization in corporate environments, it is worth noting that data visualization is on the rise in the public realm as well. For instance, The Wall Street Journal offers at its Web site a free section called Interactive Graphics, featuring many dozens of charts, maps, and other graphics. Users can select their own dimensions to build custom charts, manipulate maps, use sliders, generate mouse-over text, and more. The New York Times offers a similar section called Visualization Lab on its Web site.

Amateur investors have at their disposal a rich array of interactive charts from financial information providers and brokerage firms to analyze both individual stock performance and market trends. TD Ameritrade, for instance, offers a free visual application called Pattern Matcher that lets users pattern-match securities against certain technical market-level indicators. In effect, visualization is helping to influence stock purchases by Main Street investors.

Data visualization has even found its way into consumer communication and recreation. For
example, a free tool called Graph Your Inbox, released in September 2010, enables Gmail users to graphically view e-mail activity and trends. At about the same time, Yahoo! introduced a feature that let fantasy baseball managers view as line charts changes in standings over the season or select time periods by a number of dimensions (home runs, strikeouts, and so on). Naturally, as we become increasingly accustomed to interacting with visually presented data in our consumer lives, we come to expect the same in our business environments.

Data Visualization for Business

Graphical representations of data communicate patterns, trends, and outliers far more quickly than tables of numbers and text. With visualization, users can spot issues and problems needing attention at a glance and take appropriate action. In text-based reports and spreadsheets, the trends and issues remain hidden among dizzying arrays of numbers and text. Because of its power to communicate, data visualization is becoming more pervasive in business environments.

For instance, Airlines Reporting Corporation (ARC), a Virginia-based provider of data and transaction solutions for the travel industry, is replacing its fairly static, text-based reports with dashboards and other visual solutions to make data more accessible and easier to use for customers.

“With the graphics, it’s much easier for the eye and the brain to understand what’s happening in the data,” says Sheila Cuyjet, ARC’s director of analytic services. “It helps people see what is going to happen rather than what has already happened, and I think that’s a natural progression of analytics.”

One of the largest banks in the U.S. recently rolled out a visual dashboard to about 1,000 managers to help them monitor and manage productivity throughout the company’s regions and divisions. The MicroStrategy-based dashboard summarized information from 25 previous Excel-based reports, each with multiple pages and tabs. “Who has time to go through all that?” says the bank’s vice president of BI. “Now it’s all right in front of them.”

Data Visualization Inhibitors

Text Still Prevails Despite the growing popularity of data visualization, users still spend almost two thirds (65%) of their time analyzing data in tables and text, according to our survey. But only 12% of respondents ranked tables as “highly” useful in helping users glean insights and make decisions.

A number of sponsors of this report observed the gradual uptake of visualization. “There are still a lot of companies that don’t have visualization solutions in place, or if they do, it’s only for a small number of people,” said Niklas Derouche, CTO of BI and visualization vendor DSPanel. “But demand for visualization is growing, with the big change being a move from static to interactive to collaborative business intelligence.”

Old Habits Die Hard

Although some data is better suited to non-visual rendering, it’s clear that old habits die hard. Many executives and managers, for example, have run operations for years using text-based budgets and plans created in Excel. Likewise, business analysts, who are generally Excel jockeys, are often more comfortable with grids of data because that is how they’ve always interfaced with information.

As a result, it’s sometimes difficult to introduce visual alternatives. For example, Bruce Yen, director of business intelligence at the GUESS clothing company, encountered resistance when he replaced text- and numeral-based reports with charts. “We showed bar charts on how sales broke down by department, but some users said they could simply look at the numbers and get the same answer. Maybe the data and analysis wasn’t so complex that it needed a bar chart or some other type of visualization,” said Yen.

Poor Visual Design

Another problem is that many visual displays are so poorly designed that they actually force users to work harder to get information they need. For example, users will reject a new visual interface if it forces them to click one or more times to view related information that they think should be on a single page. Users will become frustrated if the visual design is so decorative that it is distracting or obscures the meaning of the data. Well-designed visualizations should convey much more information at a glance than users can absorb by looking at numbers.

Visual Overload

In addition, some designers may pack too much information and functionality into a visual reporting or analysis tool and overwhelm new users. The best visual displays introduce new functionality and information gradually over time. As users become more familiar with the new
environment, they typically want to view more data on a single screen and request more functions to manipulate the data. Visualization tools that expose data and functionality on demand will have higher rates of adoption among users.

Veteran BI vendors have learned this lesson. For example, IBM delivered new “on demand” features in its new IBM Cognos 10 BI tool that expose additional report design features when users click on a button labeled “Do more.”

The Role of Text

Of course, there are times when text is the appropriate way to communicate data, such as listing the top 10 customers, bottom 10 products, projects at risk, or highest-performing salespeople. It is also appropriate when users need to know the exact value or amount of a metric. Graphical views that summarize performance in a dashboard often give way to more table-driven views at lower levels of detail. That’s the case at Wyndham Exchange and Rentals (WER), a New Jersey–based vacation rental and exchange network that provides a dashboard to about 30 account managers in its inventory management team.

“Users can see something in a higher-level report, and drill down with the OLAP tool to see exactly what’s driving a decline in a certain area,” said Jeremy TerBush, WER VP of Global Analytics. “It doesn’t take them too long before they just want to see a table with data.”

Benefits of Data Visualization

Given the prevalence of text-based graphics, it’s clear that data visualization is in the early stages of industry adoption, but its popularity is growing quickly. That’s largely because it’s so easy to see (literally) and experience the benefits of data visualization.

TDWI research shows that data visualization significantly improves business insights and user productivity (that is, accelerates time to insight) and increases user adoption of BI tools. These benefits help explain the growing popularity of data visualization in corporate environments.

Business Insights

For example, nearly three-quarters (74%) of respondents to our survey rated the influence of data visualization on business insights as “very high” or “high”; 23% characterized the improvements as “moderate.”

“The combination of data discovery and visualization enables users to uncover hidden relationships they didn’t know existed,” said Judy Doherty, director of Information Management Systems at Dartmouth College in New Hampshire, which uses dashboards created with ADVIZOR Solutions to support fund-raising and alumni relations. “I’ll often hear, ‘Why didn’t we have this data before?’” Doherty said. “In fact, they did have it, in a report—they just didn’t see it.”

SWBC, a San Antonio, Texas–based financial services provider, recently replaced paper-based reports for 70 managers with a dashboard that gives account managers visibility into the profitability of their accounts, said Tommy Meuth, VP of business intelligence at SWBC. “The dashboard provides managers with instant insights about how they can decrease their loss-ratio exposure, and that will increase our profitability.”

Want to read more? Download the rest of the whitepaper!