PHD PROPOSAL: DEVELOPMENT OF AN INTELLIGENT ANALYTICS-BASED MODEL FOR ANALYZING CONSUMER BEHAVIOUR TO PREDICT THEIR ONLINE PURCHASE INTENTION

Author: Iloka Benneth Chiemelie
Published: 4th August 2019

Introduction

Based on some estimated, it is believed that Walmart gathered about 2.5 petabytes (1 petabyte = 1,000,000 gigabytes) of information per hour with these information related to consumer behaviour, transactions, devices used by the consumers and their locations (McAfee et al., 2012). It is also estimated by a Gartner IT analyst that there will be about 20 Billion (13.5 Billion in the consumer sector) that will be connected in the “Internet of Things”. If such connections are made, one can only imagine the volume of data that these devices will generate (Gartner, 2015). Such imagination can be further extended to the day that retailing data (both online and offline) can be used to provide a complete view of the consumer buying behaviour, and such can even be made better if these data are linked to all levels of the individual consumer in order to actually create “true” lifetime value calculations for the individual customers (Gupta et al., 2006; Venkatesan and Kumar, 2004). Lets further imagine the data where the data that were believed to exit in online retailing, for instance, the consumer path data (Hui, Fader, Bradlow 2009), are made available inside the store due to RFID and other technologies that are based on GPS-tracking. This can even be made further interesting if one is to visualize a day in which the online and offline integration are being run in such a way that they offer exogenous variations that would make it possible to have casual inference on retailing/marketing topics like the efficacy of advertising, email, coupons, and so on (Anderson and Simester, 2003). There could also be a day when it won’t just be possible to gather eye-tracking data from Tobii-enhanced monitors in laboratory, but such data becomes possible to gather in the field because of retinal scanning devices that become a common place among marketers (Lans, Pieters, and Wedel et al, 2008; Chandon et al, 2008).

Although all these sources might sound futuristic, the fact is that they all exist at present (although not ubiquitous, but at least albeit) and it is expected that they will soon become part of the information being utilized by marketing scientists for understanding customer-level information and optimizing firm-level objectives. It can be heuristically and simply put that, these sources of data will be creating more ‘columns’ in databases (and this columns will continue to increase) up to the extent that they provide increased ability of businesses to predict the behaviour of consumers and the influence that marketing has on such behaviour. If that becomes possible, add it to the technology (in the form of IP address tracking, registered-user-log-in, cookie tracking, loyalty card usage, and a host of others) which makes it possible for companies to gather information millions of customers, at any given time, which is linked to each and every transaction that these consumers make, linked to each and every touch point in firm-level analysis, and also linked to different platforms used for distributing goods and services, and one will see that what becomes of it is the big data that is frequently featured in the press today.

Although the lure (and lore) that comes with big data makes it tempting, the position of this paper is that the revolution of big data (McAfee et al. 2012) is actually a “better data” revolution, and this is even more so in the context of retailing. Thus, the intention of this research paper is to offer vivid description on the latest forms of data (which is the “new columns”) that are now obtainable in the retailing industry; the importance of experimentation and exogenous variations (“better columns”); in order to offer description on why machine learning and data mining (notwithstanding the pros that come with them) would never obviate the need for economic/marketing theory (which is “which part of the data to look into”); describing the statistical methods and managerial knowledge that would be used to create smart data compressions (“the columns” and their summaries) that would make it possible for the data to be aggregated by researchers; how the data can be better feed into predictive models (for instance, choice models, diffusions, CLV etc.); and finally, how likely the firms are to utilize the data for decision making. This model, which includes both the bucket and order of data, is a visual representation of the definition of business analytics offered by INFORMS (www.informs.org) to include: descriptive analytics, predictive analytics and prescriptive analytics.

Problem statement

An excellent analytics on the present, past and future of marketing analytics was offered by Li and Kannan (2016) and Little and Rubin (2014). Discussion in this work feature how marketing analytics will be used to shape future decisions making by managers when it comes to allocation of marketing mix, customer relationship management, customer privacy, personalization and security issues. Although this discussions support the idea of business analytics being important, just like most of the researchers in this area, efforts were not made in terms of developing models for how the analytics can be undertaken and this is the gap that the present study aims to fill. Mainly, the focus of application of Big Data in marketing has been on: (a) assessing the preference of consumers (e.g., Jacobs et al., 2016), (b) predicting what consumers are most likely to purchase (e.g., Ghose et al., 2012; Lu et al., 2016; Linden et al., 2003), (c) enhancing targeted advertising (e.g., Hauser et al., 2009; Trusov et al., 2016), (d) understanding consumers’ perception about brands (e.g., Culotta and Cutler, 2016; Tirunillai and Tellis, 2014), and (e), describing the competitive sphere of a market or product (e.g., Netzer et al. 2012). This research basically aims to develop an intelligent analytics-based model that can be used to analyze consumer behaviour in order to predict their online purchase intention, thereby, putting the existing theories in this context into practice.

Research objectives

In view of the discussion above, the objectives of this research are:

1.      To present a comprehensive analysis of the past, present and future of big data analytics.

2.      To discuss the impact of data analytics on effective and efficient corporate-level decision making.

3.      To develop an intelligent analytics-based model that can be used for analyzing consumer behaviour and predicting their purchase intention in the online setting.

Research framework

Figure 1: research framework

From the figure (1) above, there are two variables in this research. The independent variable is data mining and analytics, while the dependent variable is consumers’ purchasing behaviour. Basically, this research is exploratory and it will be designed to analyze the cause-and-effect relationship between the independent and dependent variables. In the case of such analysis, discussions on the independent variable will look at data: volume, value, variety, veracity, and velocity; while discussion on the dependent variable will consider how such can be used to predict intention of online consumers in relation to: purchase, repurchase, attachment, loyalty, and recommendation.

Methodology

As stated earlier, this will be an exploratory research. In the course of attaining the objectives of this paper, two steps will be undertaken. The first step will be secondary data analysis. In this case, the research will source and analyze existing studies in the context of this research in order to provide vivid understanding of the variables in this study. This will be followed by the second step, which will be development and demonstration of the intelligent analytics-based model for analyzing consumer behaviour and predicting purchase intention of online customers. In this section, the model development will be based on information from the secondary data, while the testing of developed model will be done with PYTHON Programming. For the testing stage, existing consumer data will be run in the programming software and path analysis conducted to demonstrate how data analytics can be used to predict the future purchase intention of consumers based on analysis of their past and present behaviour.

References

Anderson, E., T. and Simester, D. (2003) Effects of $9 Price Endings on Retail Sales: Evidence from Field Experiments. Quantitative Marketing and Economics. 1(1): 93-110.

Culotta, A. and Cutler, J. (2016). Mining brand perceptions from twitter social networks. Marketing Science, 35, pp. 43-362.

Ghose, A., Ipeirotis, P., G. and Li, B. (2012). Designing ranking systems for hotels on travel search engines by mining user-generated and crowdsourced content. Marketing Science, 31, pp. 493-520.

Gupta, S., Hanssens, D., Hardie, B., Kahn, W., Kumar, V., Lin, N., Ravishanker, N. and Sriram, S. (2006) "Modelling customer lifetime value." Journal of service research, 9(2), pp. 139-155.

Hauser, J., R., Urban, G., L. and Liberali, G. (2009). Braun M: Website morphing. Marketing Science, 28, pp. 202-223.

Hui, S., K., Peter, S., F. and Bradlow, E., T. (2009) "Path data in marketing: An integrative framework and prospectus for model building." Marketing Science, 28(2), pp. 320-335.

Jacobs, B., J., D., Donkers, B. and Fok, D. (2016). Model-based purchase predictions for large assortments. Marketing Science, 35, pp. 389-404.

Li, H. and Kannan, P., K. (2014) "Attributing conversions in a multichannel online marketing environment: An empirical model and a field experiment." Journal of Marketing Research, 51(1), pp. 40-56.

Linden, G., Smith, B. and York, J. (2002). Amazon.com recommendations: item-to-item collaborative filtering. IEEE, 7, pp. 76-80.

Little, R. J., & Rubin, D. B. (2014). Statistical analysis with missing data. John Wiley & Sons.

Lu, S., Xiao, L. and Ding, M. (2016). A video-based automated recommender (VAR) system for garments. Marketing Science, 35, pp. 484-510.

McAfee, A., Brynjolfsson, E., Davenport, T., H., Patil, D., J. and Barton, D. (2012) "Big data: The management revolution,” Harvard Bus Rev, 90 (10), pp. 61-67.

Netzer, O., Feldman, R., Goldenberg, J. and Fresko, M. (2012). Mine your own business: market-structure surveillance through text mining. Marketing Science, 31, pp. 521-543.

Tirunillai, S. and Tellis, G., J. (2014). Mining marketing meaning from online chatter: strategic brand analysis of big data using latent dirichlet allocation. Journal of Marketing Research, 51, pp. 463-479.

Trusov, M., Ma, L. and Jamal, Z. (2016).Crumbs of the cookie: user profiling in customer-base analysis and behavioral targeting. Marketing Science, 35, pp. 405-426.

Van der Lans, R., Pieters, R. and Wedel, M. (2008) "Research Note-Competitive Brand Salience." Marketing Science, 27(5), pp. 922-931.

Venkatesan, R. and Kumar, V. (2004) A Customer Lifetime Value Framework for Customer Selection and Resource Allocation Strategy. Journal of Marketing: October 2004, Vol. 68, No. 4, pp. 106-125.