E-Commerce, Chat Conversion Funnel, Hot Lead, AOV, Cart Load, Agent


Every retail web site is actively seeking out new innovations and approaches that create competitive advantage and increase the profitability. In general, retailers constantly monitor the behaviour of the real shoppers on the website and any changes in the market requirements. This paper presents a chat invitation web funnel structure, profiling web visitors and selection of hot leads for retail business processes through scoring method using geographic region, product page and other factors. Choosing the right hot prospects through rule base real time chat invitation method based on product type, time on page, cart load, search behavior, cookie information etc. and providing chat to those hot prospects is a special merit to this work. Active rules selection process is done using rule effectiveness indicator and chat load contribution which ensures sales revenue, chat volume and profit margin. An indirect increase in customer delight for interacting with representatives is also expected.

Full Text : PDF

  1. Agresti, A. (2002), Categorical Data Analysis, 2nd edition, John Wiley, New York.
  2.    Breslow, N. (1982), Covariance adjustment of relative-risk estimates in matched studies,
  3.   Biometrics, Vol. 38, pp. 661-672.
  4.   Constantinides E. (2004), ‘Influencing the online consumer’s behaviour: the web experience’,
  5.  Internet Research, Vol. 14 No. 2, pp. 111-126.
  6.  Dellmann, F., Kfm, D., Wulff, H., Betriebsw, D. Schmitz, S. and  Wirtschaft, F. (2003), 
  7.   ‘Statistical analysis of web log files of a German Automobile Producer findings from  practical   project concerning web usage mining’, in: Xindong Wu, Alex Tuzhilin, Jude
  8.  Shavlik (Eds.): Third IEEE International Conference on Data Mining, Melbourne, Florida, 
  9.  IEEE Press, pp. 715-718. 
  10.  E-Metrics - Business Metrics for the New Economy, available at:  
  11. (accessed July 2009).
  12.   Goldfarb, A. and Qiang, L. (2006), ‘Household-Specific Regression Using Clickstream Data’,
  13.   Statistical Science, Vol. 21 No. 2, pp. 217-255.
  14.    Griffin, N. L. and Lewis F. D. (2010), ‘A Rule-Based Inference Engine which is optimal and
  15.   VLSI implementable’, available at:
  16.   engine.pdf(accessed July 16, 2010)
  17.   Hastie, T., Tibshrani, R. and Friedman, J. (2001), Elements of Statistical Learning, Springer-
  18.   Verlag, New York.Jarvenppa, S. L. and Todd, P. A. (1996), ‘Consumer reactions to electronic 
  19.   shopping on the World Wide Web’, International Journal of Electronic Commerce, Vol. 1 No. 2,   pp. 59-88.
  20.   King, B. Andrew. (2008), Website Optimization, O’Reilly Media Inc.
  21.   Kotler, P. and Armstrong, G. (2001), Principles of Marketing, Prentice-Hall, Englewood Cliffs,         NJ.
  22.  Kucukyilmaz, T., Cambazoglu B. B., Aykanat, C. And Can, F. (2008), ‘Chat mining:      Predicting user and message attributes in computer-mediated communication’, Information  Processing and Management, Vol. 44, pp. 1448-1466.
  23.  Merrilees, Bill and Fry, Marie-Louise (2003), ‘E-trust: the influence of perceived interactivity
  24.  on e-retailing users’, Marketing Intelligence & Planning, Vol. 21 No 2, pp. 123-128.
  25.  Montgomery, A., Shibo, L., Kannan, S. and John, C. L. (2004), ‘Modelling online browsing
  26.  and path analysis using Clickstream data’, working paper, Graduate school of industrial
  27.  administration, Carnegie Mellon University, (accessed July 2009).
  28.  O’Cass, A. and Fenech, T. (2003), ‘Web retailing adoption: exploring the nature of internet
  29.  users Web retailing behaviour’, Journal of Retailing and Consumer Services, Vol. 10 No. 2, pp. 81-94.
  30.  Pallis, G., Angelis, L. and Vakali, A. (2007), ‘Validation and interpretation of Web users’     sessions  clusters’,Information Processing and Management, Vol. 43, pp. 1348–1367.
  31.  Ranganathan, C. and Ganapathy, S. (2002), ‘Key dimensions of business-to-consumer web 
  32.  sites’, Information and Management, Vol. 39, pp. 457-465.