We use cookies and other similar technologies (Cookies) to enhance your experience and to provide you with relevant content and ads. By using our website, you are agreeing to the use of Cookies. You can change your settings at any time. Cookie Policy.

E-CRM Analytics: Leveraging Data Integration for Prospective Customer Insight and Breakthrough ROI: Part 2

Originally published February 21, 2011

This is the second of a two-part series exploring the significant role of data integration in electronic customer relationship management (e-CRM) analytics. In the first part we introduced electronic customer relationship management, provided a foundation for our research, proposed our hypotheses and presented a new framework. In this second part we detail our research methodology and discuss our findings and their organizational implications.

Research Methodology

Our study was comprised of two parts. The first part was a literature review in which we looked at the results of previous studies on data integration and its inherent complexity. Based on this literature review, we designed a questionnaire to explore organizational data sources, how these data sources are integrated, the data architectures utilized for this integration, and the key integration issues facing organizations. The second part of the survey addressed e-CRM topics such as specific benefits realized, ROI expectations, user satisfaction levels and the creation or absence of a new competitive advantage. Demographic information was also collected on respondents and their organizations.

An initial survey instrument containing 17 questions was reviewed by three industry professionals to ensure appropriate and unambiguous content. The objective of this survey was to gain insight into the various data sources organizations integrate and to reveal that despite its complexity, data collection and integration for electronic customer relationship management can create value for an organization. Respondents were asked to identify benefits they had achieved or expected to achieve from their e-CRM project. Specifically, we probed responders on specific data integration procedures in their organization such as number of data sources integrated, integration of online and offline sources, integration of external data, data refresh rates and whether these data sources were centralized (in a data warehouse) or decentralized.

Additional survey questions solicited information regarding ROI, user satisfaction levels, competitive advantages and both the quantity and types of data sources integrated in respondents' e-CRM projects. These questions utilized a Likert scale to allow users to rate the success of their e-CRM initiative based on four equally weighted factors—ROI, competitive advantages, business benefits attained and user satisfaction.

Next we transformed the responses from the questions about ROI, user satisfaction, competitive advantage and benefits realized into a derived measure representing total value to an organization. We defined total value as being a combination of ROI, competitive advantage, business benefits attained and user satisfaction. The equally weighted scores obtained from these questions were added together and used as a measure of overall value generated by the e-CRM initiative. This total value figure was calculated as follows: total benefits + user satisfaction + competitive advantage + ROI = total value. The total benefits figure was calculated by summing the total number of benefits reported. This value ranged from 0 to 12. Some of the benefits reported were the increased ability to cross-sell/upsell to customers, enhanced product/service customization, increased customer retention and better customer service and inventory management. User satisfaction of the new e-CRM system ranged from 1 (not satisfied) to 7 (very satisfied). Measuring competitive advantage was calculated as the likelihood (1 = very low, 7 = almost certain) the e-CRM project enabled the company to achieve a sustainable competitive advantage. , and measuring ROI was calculated as the likelihood (1 = very low, 7 = almost certain) the e-CRM initiative generated the expected ROI.

In measuring each organization’s total value, we argue that a larger number represents more total value to the organization than a smaller number. For example, an organization with an e-CRM system that delivered eight benefits (8), created very satisfied users (7), enabled a competitive advantage (7) and delivered close to expected ROI (6) (total value = 28) would be much more valuable to an organization than a system which delivered three benefits (3), unhappy users (1), a questionable competitive advantage (1) and unsatisfactory ROI (1) (total value = 6). Using this basis for total value, we conducted statistical analyses using ANOVA to determine the correlation between our framework’s five e-CRM factors (propositions) and the total value the project created for the organization.

A website was developed for the survey and hosted at the Department of Information Systems and Operations Management at the University of North Carolina at Greensboro. A request to complete the survey was distributed to about 340 entities in the Information Systems and Operations Management Department database of organizations. This database contains data about organizations, consultants and professionals specializing in CRM technologies. A total of 115 useable responses were received and analyzed from both U.S. and international organizations, providing a 34% response rate.

Results and Discussions


Figure 1 reveals demographic information from the survey respondents. Respondents work in a wide variety of industries with the majority (49%) from CRM/technology firms and 25% from the transportation, healthcare, advertising and financial industries. Job categories for respondents range from executive management to business managers with the majority employed as CRM professionals (54%) followed by analysts (17%). Organizational revenues represent a fair mix of both small and large companies. Forty-four percent reported sales of less than a $100 million while 37% reported sales of greater than $500 million. In regard to CRM project statuses, the majority of respondents (39%) had started their CRM initiative over a year ago while 24% had started their CRM initiative less than three months ago. The majority of respondents (65%) were also clicks ’n bricks (web and store) companies while 23% were purely web retailers. In addition, 45% of respondents worked for organizations whose primary web operations were business to business (B2B) while 35% worked for organizations classified as business to consumer (B2C).

Figure 1: Respondent Demographics (N = 115)

B2B vs. B2C

Next we conducted a cross-tabular analysis to gain better insights into B2B and B2C organizations. As previously noted, 45% worked for B2B companies and 35% were from B2C companies. There was one consumer to consumer (C2C) firm while the remainders comprised the “other” category, which we presume were information-based companies such as news agencies and magazines whose revenue is primarily supported through advertising.

The survey data were analyzed across nine categories (Figure 2). The only category that showed a significant difference between the two types of firms was the data refresh rate. Forty-two percent of B2B companies refreshed their data at least once a day while 58% of B2C companies did the same. The other categories revealed very similar results when comparing the two types of firms.

Figure 2: B2B vs. B2C Analysis (N = 115)

Next we looked at the sources of data integrated by B2B and B2C firms (Figure 3). It was revealed that, in general, B2B firms integrate more data than B2C firms. The top four sources of data collected were customer demographics, online sales, offline sales, and customer communication data such as call center data, email data, etc.

Figure 3: B2B vs. B2C Data Sources (N = 115)

Next we compared the challenges and problems encountered between the firms (Figure 4). We observed that the top three problems faced by B2C firms were lack of planning, change management issues and organizational politics. The top three problems B2B firms faced were change management issues, organizational politics and lack of user buy-in. It is interesting to note that all of the top three problems of both B2B and B2C firms are organizational problems, not technical. As far as technical problems, 24% of B2B firms versus 40% of B2C firms identified data quality as a problem. In addition, 29% of B2B firms lacked user training while only 15% of B2C firms reported the same problem.

Figure 4: B2B vs. B2C Problems (N = 115)

Figure 5 displays the benefits attained between the two types of firms. In general, more B2B firms benefit from their CRM implementations than B2C firms. One interesting finding is that 51% of B2B firms reported customer service benefits while only 45% of B2C firms reported the same benefit. Apparently, B2B organizations are more effective at servicing their customers than B2C companies, which might be due to less complexity in servicing the relatively lower volumes of business customers in comparison to the much larger numbers of consumer customers. We previously reported that one of the biggest problems with B2C firms’ CRM projects was lack of planning. This problem may be a key reason behind the overall lower benefits realized in B2C firms. Although Figure 4 shows that 35% of B2B firms reported users not buying into the project, Figure 5 implies that, in general, B2B firms achieve more benefits from their CRM projects.

Figure 5: B2B vs. B2C Benefits (N = 115)

Proposition Testing
See Table 1 for ANOVA results and proposition findings. All propositions were found to be significant (p = 0.05) in their relationship to total value except for proposition 3, which proposes more total value if data is refreshed daily. A more detailed explanation and analysis of each proposition follows.

Table 1: ANOVA Results (N = 115)
Correlation of Measure vs. Total Value

For proposition 1, respondents were asked to specify the number of data sources they integrated into their data repository for the purposes of their e-CRM project. The total number of data sources integrated was calculated. Using analysis of variance (ANOVA), we determined the relationship between the total number of data sources integrated and the total value was significant (p = 0.008). This finding suggests that total value increases as organizations integrate more data sources in their e-CRM projects.

One interesting insight was that only 22% of respondents integrated all four dimensions of clickstream data, as described by Ralph Kimball and Richard Merz (2000), namely session, page, event and referrer. The session data type is a high-level diagnosis of the complete web session. Examples of segmenting web sessions by customer behavior include “Product Ordered”, “Quick Hit and Gone”, “Unhappy Visitor” or “Recent, Frequent and Intense Return Shopper” (Kimball, 2000). Referrer data identifies how the website visitor arrived at the website. A simple descriptive analysis of the percentages of different ways a visitor arrived at a website provides valuable information about how to better allocate an organization’s advertising budget. The page dimension stores data about the various attributes of each web page visited. For example, some attributes would be the page name (Product X Description, Payment Page, etc.), when it was visited, how long the user stayed on that page and where the user’s next destination was.

For proposition 2, respondents were asked whether or not they integrated offline data with their online data. Sixty-two percent said they integrated these data sources while 30% did not. The remaining 8% were unsure. Using ANOVA, we determined the relationship between those who integrated offline and online data and total value was significant (p = 0.019). Therefore, we propose that organizations that integrate both online and offline data in their e-CRM projects have significantly more benefits than organizations that do not integrate their data.

For proposition 3, respondents were asked how often they updated/refreshed the data in their data repositories. We segmented all responses into two groups – those who refreshed their data at least once a day and those who did not. Using ANOVA, we determined the relationship between frequently refreshed data (at least daily) and total value was not significant (p = 0.317). This proposition was rejected. Therefore, we propose that organizations that refresh their data at least once a day do not have a significantly higher value than organizations that refresh their data less frequently.

For proposition 4, respondents were asked whether or not they integrated external data into their central data warehouse. Seventy-four percent integrated external data in some form while 26% did not. Of those who did integrate, 62% said that external data comprised less than 20% of the total data used for analysis. Using ANOVA, we determined the relationship between integrating external data and total value was significant (p = 0.050). Therefore, we propose that organizations that integrate external data in their e-CRM projects enjoy significantly more benefits than organizations that do not integrate external data.

For proposition 5, respondents were asked to identify the data repository used for their e-CRM systems. Fifty-one percent of companies implemented legacy databases, operational data stores (ODS) or data marts as their data repositories, while 49% implemented CRM-specific databases or central data warehouses as their data repositories. Using ANOVA, we determined the relationship between the total value derived by these two segments was significant (p = 0.011). We discovered the total value derived by the group using a decentralized data repository (and not a data warehouse or a CRM-specific database) was significantly lower than the group who used a data warehouse or CRM-specific database. Therefore, we propose that organizations that implement a centralized data warehouse or CRM-specific database as their e-CRM data repository enjoy significantly more benefits than organizations that do not implement these types of data repositories.

In summary, the above propositions show that data integration is essential to accurately assessing customer needs and thus allows the firm to achieve greater e-CRM and organizational value. Therefore, we propose our e-CRM value framework (minus proposition 3—daily data refresh) is a model for generating greater total benefits and a competitive advantage for organizations engaging in e-CRM projects. To achieve the greatest amount of benefits, we suggest organizations use a data warehouse as their e-CRM data repository. This data warehouse should contain a healthy number of data sources and house all integrated data including online, offline and external data. With this architecture in place, companies are able to achieve greater profitability by obtaining a better understanding of its customers and its relationships with them.


We have presented a new e-CRM value framework to better examine the significance of integrating data from all customer touch-points with the goal of improving customer relationships and creating additional value for the firm, ultimately leading to a competitive advantage. Various issues such as the number of data sources, integrating offline, online and external data, and data architectures are discussed. We also compared and contrasted the CRM efforts of B2B versus B2C organizations and revealed some of the challenges and opportunities each face. Our findings suggest that despite the cost and complexity, data integration for e-CRM projects contributes to a better understanding of the customer and leads to higher ROI, greater number of benefits, improved user satisfaction and a greater chance of attaining a competitive advantage. Thus, when all else is equal, a company’s total value increases when a company integrates data from online, offline and external sources.

We hope that our empirical research and findings can assist practitioners and managers in identifying more efficient and effective ways of creating CRM value through data integration. It should be noted that we have only discussed the data-related issues of integration. Future research on this topic should investigate and identify managerial, financial and strategic issues that affect organizational value. In addition, other technical issues to explore include the impact of data quality and the integration role of web services.

Note: This article has been updated from the original work E-CRM Analytics: The Role of Data Integration that appeared in the July-Sept 2003 issue of the Journal of Electronic Commerce in Organizations, 1(3), Copyright 2003, Idea Group Inc.

  • Christopher BarkoChristopher Barko
    Christopher is a senior market information consultant in the Home Loans Data Management and Reporting business at Bank of America. Previously he worked for the bank as a marketing information manager in their Direct Marketing Center of Excellence. His most recent consulting engagements include website development for a leading Internet company and building affordable and robust business intelligence and reporting solutions for small businesses. His business and IT experience spans many years in a variety of consulting, business intelligence, data management, software engineering and analyst roles for a number of Fortune 500 organizations. His current interests include data mining, analytics and open source business intelligence and how these technologies improve decision making. His first book, Organizational Data Mining, a collaborative effort with Dr. Nemati, was published in 2003. His articles have also appeared in many professional and scholarly journals.
  • Ashfaaq Moosa
    Ashfaaq is a data analyst at Upromise Inc., a firm that helps families save for college. He has worked on various customer-focused data integration and data mining projects for various industries including retail, real estate, and financial.
  • Hamid NematiHamid Nemati
    Hamid is Associate Professor of Information Systems at the Information Systems and Operations Management Department of the University of North Carolina at Greensboro. Before coming to UNCG, he was on the faculty of J. Mack Robinson College of Business Administration at Georgia State University. He also has extensive professional experience as a consultant and has consulted with a number of major corporations. Dr. Nemati is the author of several books, and his most recent books are Pervasive Information Security and Privacy Developments: Trends and Advancements and Applied Cryptography for Cyber Security and Defense: Information Encryption and Cyphering. His research specialization is in the areas of decision support systems, data warehousing, data mining, knowledge management, and information security and privacy. He has presented numerous research and scholarly papers nationally and internationally and his articles have appeared in many professional and scholarly journals.


Recent articles by Christopher Barko, Ashfaaq Moosa, Hamid Nemati



Want to post a comment? Login or become a member today!

Be the first to comment!