acm-header
Sign In

Communications of the ACM

Communications of the ACM

Visualization Strategies and Tools For Enhancing Customer Relationship Management


View as: Print Mobile App ACM Digital Library Full Text (PDF) Share: Send by email Share on reddit Share on StumbleUpon Share on Hacker News Share on Tweeter Share on Facebook

In today's e-business environment, the Web site is the primary medium through which a firm interacts and performs a variety of transactions with their e-customers. In a survey conducted by General Management Strategies [1], about 80% of the responding firms indicated improving customer service as a top priority in planning their e-commerce strategies. However, unlike traditional walk-in shoppers, e-customers maintain a largely electronic relationship with an organization. Hence, businesses are attempting to exploit advances in digital technologies to foster and enhance a customer-centric culture in an electronic world. In this regard, visualization technologies and tools have attracted a great deal of attention due to their potential to enhance the customer relationshipmanagement function in organizations.

Advances in computer graphics and visual representation systems have encouraged the development of sophisticated information visualization tools, which provide capabilities for digital representation of very large data sets, innovative display methods, and superior presentation of data through an array of mechanisms such as interactive maps, 3D and multidimensional models, charts, virtual reality tools, and animations. The combination of these visualization tools with insights from data mining and data warehousing presents a plethora of possibilities for powering the customer relationship management (CRM) function within organizations. Here, we examine the key visualization strategies and tools that could be deployed in different stages of the CRM cycle. We propose a simple framework that identifies four broad areas where visualization technologies may be effectively applied and provide an overview of visualization tools available today.

A framework for applying visualization technologies in different stages of the CRM cycle is presented in Figure 1. According to the framework, the cycle consists of three stages: customer attraction, customer acquisition, and customer analysis. Customer attraction occurs when consumers are drawn to a Web site to perform certain online transactions. In this stage, visualization tools can add considerable value in displaying search results, viewing products, and in presenting product-related information. During customer acquisition, companies try to convert Web site visitors and browsers into (hopefully repeat) buyers. Visualization tools can play a key role in product evaluation and selection at this stage. Customer analysis gathers a variety of customer data that is logged and analyzed in various ways to better understand and identify customer browsing behaviors and buying patterns. Here, visualization tools play a significant role in enhancing information representation to key decision makers within the organization.

Back to Top

Visualization of Search Results

Search is one of the most common functions customers perform online. When consumers search for information, they not only expect to find the most relevant and appropriate information for their needs, but also expect to obtain this information in the shortest possible timeframe. Conventional search techniques deployed by online merchants typically require the user to search by selected key words. The results are displayed in a simplistic manner, often as a list, sorted by relevance and the type of metatags used by the site. Though there is a difference in the search algorithms used by different Web sites and search engines, the search results are presented in a similar manner. By devising innovative ways to visually present the search results to customers, Web businesses can differentiate themselves from other competitors. This makes the display of search results a potential area for application of visualization tools.


By devising innovative ways to visually present the search results to customers, Web businesses can differentiate themselves from other competitors.


State-of-the-art visualization techniques could go a long way toward making the Web site search faster, easier, and more intuitive for customers. These technques could allow for simultaneous display of a large amount of information arranged visually in such a way that it is not overwhelming for the users. Several visualization tools could be used to improve the presentation of search results. For instance, a visual search tool called Search Query Weighted Information Display (SQWID),1 developed at the Georgia Institute of Technology, provides a dynamic node-link-graph display of the search results. It identifies three key terms closely related to the search query and displays a triangle of nodes with these key terms. The Web pages matching these key terms are arranged around these nodes, the closer ones being more relevant.

VisIT,2 developed at the University of Illinois at Urbana-Champaign, deploys a metasearch on common search engines and displays the results in a search space (see Figure 2a). Web sites are represented in this space as simple icons, differentiated by multiple colors with lines and arrows drawn between the icons. Each icon represents a Web page, and each little group of icons represents similar sites. Arrows between page icons indicate references between the corresponding Web pages. A highly relevant page for the current search can be easily determined from the number of inbound arrows it has from other pages. Additionally, a pop-up box with text from the actual page is displayed when the cursor is moved on a page icon, making it easier to select a page the user seeks.

NIST Information Visualization Engine (NIRVE)3 presents a similar graphical summary of search results with icons representing clusters of Web sites, arranged on a sphere or 2D surface. Multicolored arcs are used to indicate relationships among the clusters. The user can open a cluster to view the sites contained there and can mark them as "good," "bad," or "unsure," depending on their relevance.

More recently, researchers at University of California at Berkeley developed a new search interface called FLexible information Access using Metadata in Novel Combinations (Flamenco, [4]) that allows users to both refine and expand a query, while maintaining a consistent representation of the metadata structure (see Figure 2b). This integration of free-text search with a large information space of metadata, allows the user to recursively follow links and search terms, thereby providing a continuous flow of interaction without interruptions.

There are other tools that can be incorporated as a toolbar within the Web browsers. Girafa is one such tool that presents visual previews of the Web sites (search results) without the need to visit the respective sites. The search results are displayed along with thumbnail images of the sites (see Figure 2c). Moreover, while viewing a thumbnail image, Girafa allows a consumer to view the thumbnails of other sites listed on this page. Kartoo is a similar visualization-based search tool that provides a graphic map of metasearch results (see Figure 2d). Other examples illustrating the use of visual search interfaces are Conceptspace implemented at the London Business School, and 3D Engineering Space Search (3DESS) system developed by Purdue University that searches through 3D engineering parts.4

Back to Top

Visualization in Product Display and Presentation

One of the primary drawbacks of online shopping is the customer's inability to feel or touch the physical product. This limitation can be countered by deploying visualization tactics for displaying products online. Common tactics used by online vendors include displaying product images, and showcasing products via 3D views.

Quick-Time Virtual Reality (QTVR) is a viable alternative for online merchants to present 3D views of their products. QTVR enables 360-degree views of products and allows customers to click on photos to view their products from a variety of angles. Several retailers, including Sharper Image and Lands' End, have adopted QTVR technology. Although 3D options can be more expensive than traditional image views, customer satisfaction that results in more purchases could easily offset the high costs. This technological edge can prove to be a boon for retailers in the race for market supremacy.

Visualization applications for enhancing the display and presentation of products abound on the Web. The Scientific Visualization and Digital Media Group5 developed a large-scale hypermedia regional information system called Virtual Sardinia. This project collects a large amount of heterogeneous data such as geographic, archaeological, and historical information and visually presents it in a way that allows users to take a virtual trip through the Italian island. The front end includes a detailed 3D reconstruction of Sardinia from satellite pictures. Using a high-speed, 3D scene viewer called i3D and a Spaceball, the user can navigate virtual Sardinia. The 3D terrain is subdivided into different quadrants, with each hyperlinked to a high-resolution image. Interesting tourist sites on the model are identified by 3D objects with an associated hyperlink. Throughout the trip, users have the option of viewing videotapes and clips of different parts of the island. The project also employs another software product called Web Show to browse and view long image sequences [3]. Figure

Several organizations also use visualization tools to present their products and services to e-customers. For example, Canadavirtualtours.com has developed virtual tours for several Canadian cities. One of its sites, manitobavirtualtours.com offers a virtual tour of several tourist attractions, restaurants, and bars in different locations at Manitoba. Clicking on a virtual tour for a restaurant or a bar in Manitoba gives a 3D rotary picture of the restaurant, providing a picture of its ambience and decor. Prototyping tools for preparing 3D views of products are also commercially available. For example, Quest3D is a tool enabling vendors to create VR presentations and product visualization.

Another area for potential application pertains to product self-configuration and self-customization. Product tailoring has become a common feature in the online marketplace, with firms allowing customers to configure products to suit their particular tastes and preferences. For instance, virtue3D.com uses a Virtue3D Room Designer that allows customers to choose a room they want to decorate, and furnish it completely with furniture, drapes, and accessories in a 3D visual environment. The customer can experiment with different decors, colors, furnishings, and make purchases. A similar application used by apparel retailers is the concept of the virtual dressing room or virtual model that users can customize by choosing from a menu of hairstyles, hair color, skin tones, face shape, eye color, eye shapes, and so on. The online model allows users to "try on" different garments to judge style and fit.

Back to Top

Visualization in Online Product Evaluation and Selection

One of the key elements in designing an online store is the presentation of product information to the customers to make product evaluation and selection easier. A common tactic used by Web sites is to provide customers with online catalogs that contain information on product specifications, features, prices, and other related information. In a typical online shopping environment, a customer is required to browse multiple pages of the Web catalog, gather information across these pages, and make comparisons. This process gets tedious if the customer wants to evaluate and compare products across different online stores. There are several sites and shopping bots that allow for comparison shopping across multiple online stores, but most of them use a textual display format in the form of tables. This is another potential area for applying visualization tools. A multidimensional visualization mechanism called Visual One-Page Catalog (VOPC) developed by IBM uses innovative display techniques for online product presentation, comparison, and evaluation.


Visualization tools offer new ways to slice, dice, and analyze huge amounts of customer transaction data.


VOPC uses parallel coordinates to display multiple products along with their key attributes in a single Web page. Each axis on the page represents a unique product attribute and customers can decide on the nature and number of product attributes. A customer can classify the product attributes as first, second, and third tier based on their importance in the purchase decision. Then, the customer specifies lower and upper bounds on the attributes. For instance, if customers are shopping for computers, they can use price as an attribute in an axis, and set the lower and upper price limits they are willing to pay for the computer. Once a customer specifies multiple product attributes and related bounds, product lines are drawn across all the attribute axes based on the product's values on these attributes. If a customer wishes to compare multiple products, multiple product lines are drawn across different attribute axes; VOPC displays desirable values on the right side of product attribute axes. Therefore, a customer can "see" the product lines and easily identify the most desirable alternatives. With a click on the appropriate item, the customer can receive additional information and complete the purchase. Thus, VOPC affects product information across several pages, or different sites into a single page. It provides customers with a visualized comparison of product alternatives and facilitates their decision process.

Another visual tool handy in online product selection is VizServer from Inxight Technologies. This is not an e-cataloging product per se, but a tool potentially adaptable to the e-product cataloging domain. This tool enables the compression of voluminous product information into a single page through different visual forms. In addition to such tools, a number of other commercial online product cataloging services are available. Sites such as HiShop, JordynAsp, Infrontier's Product Cataloging and Digital Snaps' Product Cataloging system offer tools that deploy visualization tactics for presenting product-related information to consumers. Elecat-reader is a more recent product cataloging tool providing page visualization facility while preserving the original layout and connecting pages with the integrating database.

Back to Top

Visualization of Customer Clicks

Most businesses are currently facing tremendous data overload due to increased online interactions with customers and other business partners. Visualization tools offer new ways to slice, dice, and analyze huge amounts of customer transaction data. These tools provide specialized views, graphs, reports, and multidimensional charts of customer data better than the conventional methods of textual, tabular, or simple graphical formats of presentation. The evolution of these visual analysis tools is in direct response to the increasingly ambitious analyses sought by managers, and the expanding size of customer data warehouses.

NIST has developed two toolsWeb Variable Instrumenter Program (WebVIP) and Visual Variable Instrumenter Program (VisVIP)for assessing the usage levels and usage patterns of Web sites. WebVIP captures different surfing parameters such as page transitions, use of buttons, time spent on a page, and so on. Using the data available from WebWIP, VisVIP generates a 2D and 3D layout of the Web site in the form of a directed graph. The graph visually represents different pages as nodes and a user's navigation pattern is represented by curves running through these nodes. Arrowheads denote the direction of the navigation and vertical bars connected to the nodes are used to represent the time spent by a user on the Web page, providing a 3D visualization of user navigation paths through a Web site.


The evolution of these tools is in direct response to the increasingly ambitious analyses sought by managers, and the expanding size of customer data warehouses. Visualization and analysis of click-stream data identifying visitors' routes through Web sites is also used to determine effectiveness of marketing and merchandising efforts at online stores. It can help the store identify shopping patterns of different groups of customers, the areas on the Web site where it loses customers, and the effectiveness of online promotions. Research on data visualization has led to several data-analytic products commercially available for businesses. For example, SPSS Inc. has developed a Java-based software development kit called nVizn that allows users to build highly visual, interactive, and customized analytical applications. Netgenesis, a customer intelligence firm specializing in measuring and analyzing a company's e-business metrics, uses nVizn to provide visually stronger analyses to their clients (see Figure 3a). Another product, Ebizinsights, developed by Visual Insights, (now known as Advizor Solutions) helps companies track the customer activities, visitor behavior, visually analyze click-stream data, and customer response to promotions. It has two key modules: the Visual Path analysis module that graphically depicts the customer's navigation paths through the company's Web site and the Visual Portal module that lets managers tailor numerous graphical reports concerning the customer's use of the Web site. Several such commercial visualization packages (examples include DataScope Explorer from Cygron, and Enterprise Visualization Suite from Knowledge Relay) provide a flexible environment for data visualization to analyze and resolve specific customer-related business problems. New Java-based tools simplify the analysis process, providing simple Web-based interfaces to interact with graphic representation of data. Multipanel, multipage views have become a common feature in visualization tools and such displays simplify comparison of different business parameters, reducing the time for analysis as well as decision making. These visualization-based data analytic tools can help managers interpret customer information, understand different nuances of customer relationships, and present and communicate justifiable conclusions to others.


Back to Top

Conclusion

E-commerce is taken for granted in the marketplace today, and yet it is still common to find Web sites with poorly designed search functions, clumsy online catalogs, unimpressive product displays, and limited tools to assist customers in their purchase process. Even those Web sites that fare well in these parameters are working hard to attract and retain customers in this challenging environment. Competition in the online marketplace is intense and sites that make it easy for the customer to interact with them are the ones that are going to be successful over a longer term. In this regard, visualization is proving to be a strong mechanism for attracting, assisting, acquiring, and analyzing online customers.

Visualization tools provide effective mechanisms for understanding customer behavior, increasing customer loyalty, and providing effective customer serviceall of which will help a site to better cater to its customers. In this article, we have presented how visualization tools can be effectively applied to improve and enhance the different stages of the CRM function. We have also outlined a number of tools and illustrations to serve as a pointer to IT managers for deploying visualization technologies in their organizations. Meaningful visualization applications aid customer decision making, which could lead to a brighter future for many firms trying to build a thriving business online.

Back to Top

References

1. Butler, S. B2B and CRM: Managing Customer Relationships (Dec. 15, 2000); www.emarketer.com/analysis/ecommerce_b2b/20001215_ecrm.html.

2. Eick, S G. Visualizing online activity. Commun. ACM 44, 8 (Aug. 2001), 4548.

3. Gobetti, E. and Leone, A.O. Virtual Sardinia: A large-scale hypermedia regional information system. Computer Networks and ISDN Systems 28 (Sept. 1996), 711; www.crs4.it/PRJ/VIRTSARD/.

4. Hearst, M., English, J., Sinha R., Swearingen, K., and Yee, P. Finding the flow in Web site search. Commun. ACM 45, 9 (Sept. 2002) 4249.

Back to Top

Authors

Shobha Ganapathy (shobha_gan@yahoo.com) is a senior marketing analyst at FedEx Services, Memphis, TN.

C. Ranganathan (ranga@uic.edu) is an assistant professor at the Department of Information & Decision Sciences at the University of Illinois at Chicago.

Balaji Sankaranarayanan (bsankara@indiana.edu) is a doctoral student in information systems at Indiana University, Bloomington, IN.

Back to Top

Footnotes

1See www.cc.gatech.edu/grads/m/Scott.McCrickard/sqwid for more information.

2See www.visit.uiuc.edu/ for more information.

3See www.itl.nist.gov/iaui/vvrg/cugini/uicd/nirve-home.html for more information.

4See conceptspace.london.edu for more information on Conceptspace. For more on 3DESS, see tools.ecn.purdue.edu/~cise/dess.html.

5See www.crs4.it/~zip/group_homepage.html for more information.

Back to Top

Figures

F1Figure 1. Applying visualization technologies in CRM cycle.

F2Figure 2. Search-related visualization tools.

F3Figure 3. Visualization-based data-analysis tools. A snapshot from nVizn.

UF1Figure. Snapshots from Virtual Sardinia

Back to top


©2004 ACM  0001-0782/04/1100  $5.00

Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.

The Digital Library is published by the Association for Computing Machinery. Copyright © 2004 ACM, Inc.


 

No entries found