The Role of Data Mining and Fusion in Modern Marketing Management - Article Example

Critically describe the role of data mining fusion in modern marketing management Data mining fusion refers to the sorting and selecting of information deemed relevant to the individual / organisation who has undertaken the activity. Frawley et al define it as; “The nontrivial extraction of implicit, previously unknown, and potentially useful information from data.” (Frawley et al; 1992, 213) And, Hand et al define the process as; “The science of extracting useful information from large data sets or databases.” (Hand et al; 2001) Whilst there are both advantages and disadvantages with the technique, generally speaking data mining / fusion, which can also be known as knowledge discovery in databases, can prove to be a very valuable tool in the modern marketing manager’s toolbox. Managers seek sustainability. As managers of a business we aim to achieve high profit margins but we don’t want to be a ‘fly by night’ organisation, which is all too common in hospitality. It is agreed by all that we are in it for the long haul, we as managers are seeking sustainability. Strategy is extremely important, looking to the long-term, to be proactive rather than constantly work at the operational and tactical levels, constantly having to ‘fire fight’ to survive in a harsh marketing environment. Hence, information is essential and needed so as to permit marketing management to carry out marketing analysis, planning, implementation and control. “To manage a business well is to manage its future and to manage the future is to manage information.” (Harper: 1961) (cited in Kotler, Armstrong Saunders & Wong: 1999, 316) “…marketing managers need information. They need information about customers, competitors, and other forces in the marketplace. In recent years, many factors have increased the need for more detailed information. As firms have become national and international in scope, the need for information on larger and more distant markets has increased. As consumers have become more affluent and sophisticated, marketing managers need better information on how consumers will respond to products and other market offerings. As competition has become intense, managers need information of their marketing tools. Because the environment is changing rapidly, marketing managers need more timely information.” (Erdem & Swift: 1998) Data mining / fusion with its ability to search for hidden relationships, patterns, correlations and interdependencies is valuable to marketing managers in relation to marketing research (MR) and Marketing Information Systems (MIS). MR and MIS allow the marketing manager to set about achieving their key goal; to identify, anticipate and meet (or exceed) at a profit, the desired requirements of the consumer. To a great extent the benefits of data mining and degree of its success will largely depend on how the information is used, the purpose of its need and skill of the marketer / marketing team using the information. The size and type of organisation involved is also important regarding the relevance of data mining as also is the product offering; whether the company offers only one product or if it is a multi faceted company. A small one product offering company for example may find it more beneficial, especially in terms of cost, to buy a set of data, relevant to its needs, from the data warehouse of a specialist research company. This information obtained would then be fed into its own MIS and used to market perhaps a new product offering or perhaps to tailor an existing product to fulfil a market niche. This point along with the advantages of the data mining technique of appending data to third party files is highlighted by Kathy Seal in relation to the hotel industry; “In a highly competitive marketplace, to attract and retain customers new tools are needed to sort, shift and analyze the daily compilation of guest data that occurs at property level. For hoteliers, it is necessary to plume this data in order to make sense of it, or to even justify the process of collecting it in the first place… One technique is to use guest data is to append it to ‘third party files,’ such as demographic or psychographic information from a data house, which has researched consumer habits according to zip code. For example, a marketing director who knows there will be a NASCAR race in his city can pull guest zip codes out of his data warehouse, and using information bought from the data house, find out which guests live within driving distance of the hotel and are likely to enjoy race car driving… Then he can direct mail them information about a NASCAR weekend package.” (Seal: 1998, 40) Similarly another idea was to use MR and MIS to identify couples who honeymooned in the hotel and to contact them with an anniversary weekend package. This usage is clearly for promotional purposes and may create extra revenue, it may be particularly advantageous if analysis had identified a problem with weekend occupancy rates, but how strategic is this use? Is it more problem solving and short term rather than long term and sustainable? Jim Bowen (1998) further highlights the important role that data mining / fusion can play with regards to market segmentation and subsequently positioning and targeting. This is especially important as Myers states: “One of the most important strategic concepts contributed by the marketing discipline to business firms and other types of organisation is that of market segmentation” (Myers; 1996) An example of how data mining in specific areas and benefits of subsequent segmentation are evident in geodemographic segmentation such as PRIZM. This is an American example which is similar to the previous NASCAR example in that profiles are created according to different zip codes. The general premise is that ‘birds of a feather flock together.’ The example given by Bowen clearly indicates how the marketer can take advantage of segmented information with regards to targeting and product positioning; “… the worlds largest users of scheduled airlines belong to suburbs called the Urban Gold Coast. The suburbs are comprised of upscale high-rise neighbourhoods in only a handful of big cities. Urban Gold Coast tops many demographic lists; most densely populated, most employed, most white-collar workers, most renters, most childless and most New York based. Almost two thirds live in residences worth more than $200 000; decorating their home according to Metropolitan Home, buying their clothes at Brook Brothers; and frequenting the same hand starch Chinese laundries. In Urban Gold Coast, residents have the lowest incidence of auto ownership in the nation; these cliff dwellers get around by taxi and rental car. Urban Gold Coast residents usually eat our for dinner and lunch” (Bowen; 1998, 289) Clearly for the savvy modern marketing manager the information gathered here from data mining could have very advantageous outcomes. This is especially so if a number of companies were to perhaps form a strategic alliance of mutual benefit, for example; an alliance between certain restaurants and taxi firms could prove very lucrative. The opportunities could be endless. In contrast perhaps to some smaller organisations, larger organisations may prefer and have the ability in terms of time, finance and human resources to carry out their own research as necessitated and use it to their advantage. The Days Inns Hotel Group in America used a combination of quantitative and qualitative descriptive research to answer the question ‘what does the customer want?’ so as to assist in better strategic decision making regarding media selection, geographic coverage, promotion and also in assuring customer satisfaction. Hotels were rated from a guests’ point of view by means of scoring various attributes (divided into two groups involving employee attitude and efficiency, and, hotel ambience and comfort). The results were then used in human resource strategies such as in training and development initiatives / policies; and also in financial strategy, in allocating funds for initiatives / policies in design / construction. Results were also used in future location considerations. (Atkinson: 1988) Information gathered was also: “…consolidated into a marketing strategy in which the marketing department selected media placements that would inform the guests of the company’s intent and desire to meet their needs and expectations. One of the marketing strategies was called ‘zero defects.’ Stated in simple language, it means that each hotel focuses on making sure everything in every room works properly.” (Atkinson: 1988, 31) Also in the hotel sector in the U.S. The Market Metrix Hospitality Index (MMHI), created in the 1990’s by the Cornell Hotel and Restaurant Administration Quarterly, is the industries largest in-depth measure of hotel performance based on guest evaluation. This index highlights how MR and MIS have been successfully used in the hotel industry. Although not specifically classified as a MIS, when used by individual hotels with their own research and market intelligence / environmental scanning data, can form the basis of a MIS. Due to the ‘emotion’ associated to guest evaluation and differing interpretations the information from the index should not be used alone as a basis for strategic management decisions. However, it has proved to be a very useful tool; it spots industry trends such as the dip in customer satisfaction after 9/11. “The Market Metrix research has demonstrated that emotions play an important role in hotel customers’ satisfaction and loyalty, and those emotions are a better predictor of customer loyalty than are traditional measures of product and service satisfaction. Guests are willing to pay substantially more per night for the promise of experiencing certain emotions during their stay.” (Barsky & Nash: 2003,174) The index found that business travellers would pay more for comfort and be less price sensitive. Investing in comfort would thus increase loyalty, the problem is however, defining comfort, is it an emotion created due to the product, that is, the actual aesthetics – furniture and design, or, is it created due to the service provided by staff. Independent research by individual hotels will be required to establish the specifics of their own customers needs, thus the decision to for example introduce refurbishment or customer service training (perhaps even both) may be made. Gargano and Raggad (1999) acknowledge the point that Berson and Smith (1997) highlight; that there is a great synergy between data mining / fusion and data warehousing; “The synergy created between the holistic paradigms of data warehousing and data mining allows goal orientated decision makers to leverage their massive data assets, thereby improving the effectiveness and quality of their decisions.” (Gargano and Raggad; 1999) For industries such as the hotel industry, or organisations carrying out their own research, establishing their own data warehouse and using data mining / fusion techniques to fit their own specific purposes; there are a number of data mining models and tools available. Garango and Raggad (1999) highlight that; “The tools used in data mining are simple, concise, easy to implement algorithms that model non-random (i.e. statistically significant) relationships (or patterns) in large historic data sets. These models can then be applied to novel data in order to classify, predict associate or optimize.” (Gargano and Raggad; 1999) Garango and Raggad (1999) also note that there are a wide range of tools inspired by different paradigms including; artificial intelligence, fuzzy logic, decision trees, rule induction methods, genetic algorithms, artificial neural networks, associative memories and cluster techniques. Each tool or model will approach the data from a different perspective but will share the same common characteristic that; “Each adaptively improves as the model learns from the data set and more knowledge is acquired.” (Gargano and Raggad; 1999) With regards to using data mining / fusion models and techniques one factor which may be deemed as a drawback may be the time required for training and the associated cost. “Data mining models have a training phase, when the model learns patterns and relationships hidden an historical data set, followed by an implementation phase, then the model is actually used.” (Gargano and Raggad; 1999) However, a long-term advantage and general consensus is that once trained models quickly produce results and that; “The amount of technical skill required to implement and integrate these models into the decision making process is typically minimal.” (Gargano and Raggad; 1999) Marketing managers must carefully consider whether the benefits of this time and expenditure are justified / viable in relevance to the overall strategic gain as a result. The type of research and data required must also be carefully considered, qualitative, quantitative or a combination of both. It is important that marketing managers do not get completely absorbed within the complexities and technicalities of the information process and lose sight of the overall objective(s). Such models, tools and techniques are there as an aid but the definitive qualities of good managers themselves must not be forgotten or sidetracked as evident by the KFC example. In 1985 in the post communism era, the Chinese government approached KFC in a bid to bring commercial business and new technology, namely fast food processing systems so as to upgrade Chinese food production, to the new Republic. KFC carried out exploratory research and had the assistance of some limited government economic data. There was, however, a distinct lack of information and the government did not release data on the restaurant industry. Research would have to be carried out on Chinese culture and eating trends, the Chinese had not been subject to MR before and thus it was difficult to know how they would react. However, a huge market waited if KFC took up the opportunity presented, the government had offered prime sites in major cities. Despite lack of information KFC decided to take the risk and pursue the opportunity. This was an excellent move on the company’s part, leading to further penetration of the Pacific Rim with approximately 2500 restaurants in 2000. (Malhotra: 1999) This success story highlights that MR and MIS are important for management when making strategic decisions, however, it should not stand alone but used in conjunction with the intuition of experienced hospitality managers: “Marketing research does not replace management skills, it reinforces management thinking by testing conventional wisdom and commonly held assumptions in business. In short, marketing research provides new perspectives in strategic planning.” (Tan Tsu Wee: 2001, 246) Marketing managers do not need to be inundated with research information to the point where they would become obsessed by it and ignore their own inclinations and experienced knowledge when making strategic decisions. However, as highlighted by Brown and Kros (2003), missing data can also hinder the overall marketing process. “The actual data mining process deals significantly with prediction, estimation, classification, pattern recognition and the development of association rules. Therefore, the significance of the analysis depends heavily on the accuracy of the data base and on the chosen sample data to be used for model training and testing. Data mining is based upon searching the concatenation of multiple databases along with a variable percentage of inaccurate data, pollution and noise. The issue of missing data must be addressed since ignoring this problem can introduce bias into the models being evaluated and lead to inaccurate data mining conclusions.” (Brown & Kros: 2003,1) Data mining / fusion plays a vital role regarding customer relationship marketing (CRM). The retail industry has been at the forefront of data mining with much noted success. Many UK retailers acknowledge the data mining benefits of a consumer database. Hanford’s and Sainsbury’s use Bram Viper software whereas Tesco and John Lewis use Dunn Humby. Successes in retailing include the ‘nappies-beer’ link on Friday evenings as acknowledged by Wal Mart in the US. Placing the both products side by side meant that fathers bought more beer than intended when they went to buy nappies after work. (Dennis, Marsland and Cockett: 2001) Similarly in the UK Woolworths claim to have boosted women’s toiletries sales by £5 million plus, after introducing a similar system to the cost of £2 million and they firmly believe that; “Incorporating data mining and customer database aspects within a framework of knowledge management can help increase knowledge value.” (Dennis, Marsland and Cockett: 2001; 369) Data mining can prove useful with regards to customer loyalty schemes and beneficial to catalogue retailers regarding what types of discounts / offers to be used with regards to specific customers. However, one factor which marketers must always be careful in addressing is that of privacy and data protection. Evidently there are many examples of how varying organisations use data mining / fusion techniques to assist them in their marketing plans and in striving towards achieving their overall specific strategic objectives. Data mining clearly has both advantages and disadvantages, and the extent and method that any given company employ to extract and make use of data will largely depend on their individual size, capacity and resources. It can be concluded that knowledge via information is vital for the marketing managers of all organisations. Data mining / fusion is of great importance for market research and marketing information systems which can be viewed as necessary for successful marketing projects. All projects should have a clear purpose and be implemented correctly. Research can be time consuming and expensive, therefore it is essential that the correct methods are used and that it is relevant. There is a danger of information overload and that management should not become totally reliant on the findings. There are a number of research methodologies available to a marketing manager, all methods are valuable in their own way but will blend with each other for better strategic decision making. (Lynn & Lynn: 2003) Research adds value to decision making: “…research that includes understanding a phenomenon as one of its objectives may potentially lead to long-term improvements in management, operating practices or competitive strategies.” (Van Scotter & Culligan: 2003, 15) Finally, if to be beneficial and to be used effectively and strategic tools, MR and MIS cannot be separated: “Marketing research and information systems are one special Siamese twins, symbiotic in nature, born together not to be tampered with in any way. Any attempt at separation would be cardinally unholy. To remain in the good grace of business visionaries and to give the organisation a differential advantage, we should keep the two together for better way of reducing the fog of uncertainties surrounding management decisions.” (Demirdjian: 2003, 225) REFERENCES Atkinson, A (1988) Answering The Eternal Question: What Does The Customer Want. Cornwll Hotel and Restaurant Administration Quarterly 29(2) pp.12 Barsky, J and Nash, L (2003) Customer Satisfaction: Applying Concepts to Industry-wide Measures. Cornwll Hotel and Restaurant Administration Quarterly 44(5/6) pp.173-184 Berson, A and Smith, S.J. (1997) Data Warehousing, Data Mining and OALP New York: McGraw Hill Bowen, J.C (1998) Market Segmentation in Hospitality Research: No Longer a Sequential Process. International Journal of Contemporary Hospitality Management 10(7) pp.289-296 Brown, M and Kros, J (2003) Data Mining and The Impact of Missing Data. Industrial Management and Data Systems 103(8) pp.1 Demirdjian Z S (2003) Marketing research and information systems: The unholy separation of Siamese twins. Journal of American Academy of Business 3 (1/2) pp.225 Dennis, C and Marsland, D and Cockett, T (2001) Data Mining for Shopping Centres – Customer Knowledge – Management Framework. Journal of Knowledge Management 5(4) pp368-374 Erdem, S.A and Swift, C (1998) Items to Consider for Just-in-Time Use in Marketing Channels: Towards development of a Decision Tool. Industrial Marketing Management 27(1) pp.21-29 Frawley, W and Piatetsky-Shapiro, G and Matheus, C (1992) Knowledge Discovery in Databases: An Overview. AI Magazine pp. 213-228 Gargano, M and Raggad, B (1999) Data Mining a Powerful Information Gathering Tool. OCLC Systems and Services 15(2) pp.81-90 Hand, D and Mannila, H and Smyth, P (2001) Principles of Data Mining. MIT Press, Cambridge, MA: MIT Press Harper, M (1961) A New Profession to Aid Management. Journal of Marketing pp.1. In Kotler, P and Armstrong, G and Saunders, J and Wong, V (1999) Principles of Marketing 8th New Jersey: Prentice Hall Kotler, P and Armstrong, G and Saunders, J and Wong, V (1999) Principles of Marketing 8th New Jersey: Prentice Hall Lynn, A and Lynn, M (2003) Experiments and Quasi-Experiments: Methods for Evaluating Marketing Options. Cornwll Hotel and Restaurant Administration Quarterly 44(2) pp.75 Malhorta, N (1999) Marketing Research: An Applied Orientation 3rd New Jersey: Prentice Hall Myers, C.S (1996) Trust Commitment and Values Shared in Long-Term Relationships in the Services Marketing Industry. University of Nevada, Las Vegas Masters Thesis. In Bowen, J.C (1998) Market Segmentation in Hospitality Research: No Longer a Sequential Process. International Journal of Contemporary Hospitality Management 10(7) pp.289-296 Seal, K (1998) Data Management Helps Marketing Efforts. Hotel and Motel Management 213(3) pp.40-42 Tan Tsu Wee, T (2001) The use of Marketing Research and Intelligence in Strategic Planning: Key Issues and Future Trends. Marketing Intelligence and Planning 19(4) pp. 245-254 Van Scotter, J and Culligan P.E. (2003) The Value of Theoretical Research and Applied Research For The Hospitality Industry. Cornwll Hotel and Restaurant Administration Quarterly 44(2) pp.14 BIBLIOGRAPHY Aaker, D and Day, G (1980) Increasing the Effectiveness of Marketing Research. California Management Review 23 (2) Atkinson, A (1988) Answering The Eternal Question: What Does The Customer Want. Cornwll Hotel and Restaurant Administration Quarterly 29(2) pp.12 Barsky, J and Nash, L (2003) Customer Satisfaction: Applying Concepts to Industry-wide Measures. Cornwll Hotel and Restaurant Administration Quarterly 44(5/6) pp.173-184 Berson, A and Smith, S.J. (1997) Data Warehousing, Data Mining and OALP New York: McGraw Hill Bowen, J.C (1998) Market Segmentation in Hospitality Research: No Longer a Sequential Process. International Journal of Contemporary Hospitality Management 10(7) pp.289-296 Brown, M and Kros, J (2003) Data Mining and The Impact of Missing Data. Industrial Management and Data Systems 103(8) pp.1 Chisnall, P. (2001) Marketing Research. 6th edition. Berkshire: McGraw Hill Publishing Company Demirdjian Z S (2003) Marketing research and information systems: The unholy separation of Siamese twins. Journal of American Academy of Business 3 (1/2) pp.225 Dennis, C and Marsland, D and Cockett, T (2001) Data Mining for Shopping Centres – Customer Knowledge – Management Framework. Journal of Knowledge Management 5(4) pp368-374 Erdem, S.A and Swift, C (1998) Items to Consider for Just-in-Time Use in Marketing Channels: Towards development of a Decision Tool. Industrial Marketing Management 27(1) pp.21-29 Frawley, W and Piatetsky-Shapiro, G and Matheus, C (1992) Knowledge Discovery in Databases: An Overview. AI Magazine pp. 213-228 Gargano, M and Raggad, B (1999) Data Mining a Powerful Information Gathering Tool. OCLC Systems and Services 15(2) pp.81-90 Hand, D and Mannila, H and Smyth, P (2001) Principles of Data Mining. MIT Press, Cambridge, MA: MIT Press Harper, M (1961) A New Profession to Aid Management. Journal of Marketing pp.1. In Kotler, P and Armstrong, G and Saunders, J and Wong, V (1999) Principles of Marketing 8th New Jersey: Prentice Hall Hines T (1995) Management Information for Marketing and Sales: The Marketing Series: Chartered Institute of Marketing Oxford: Butterworth Heinemann Jobber, D and Rainbow, C (1977) A Study of the Development and Implementation of Marketing Information Systems in British Industry. Journal of the Marketing Research Society 19 (3) pp.104-111 In Jobber D (2001) Principles and Practice of Marketing 3rd London: McGraw Hill Jobber, D (2001) Principles and Practice of Marketing 3rd London: McGraw Hill Kotler, P and Armstrong, G and Saunders, J and Wong, V (1999) Principles of Marketing 8th New Jersey: Prentice Hall Lynn, A and Lynn, M (2003) Experiments and Quasi-Experiments: Methods for Evaluating Marketing Options. Cornwll Hotel and Restaurant Administration Quarterly 44(2) pp.75 Malhorta, N (1999) Marketing Research: An Applied Orientation 3rd New Jersey: Prentice Hall Myers, C.S (1996) Trust Commitment and Values Shared in Long-Term Relationships in the Services Marketing Industry. University of Nevada, Las Vegas Masters Thesis. In Bowen, J.C (1998) Market Segmentation in Hospitality Research: No Longer a Sequential Process. International Journal of Contemporary Hospitality Management 10(7) pp.289-296 Seal, K (1998) Data Management Helps Marketing Efforts. Hotel and Motel Management 213(3) pp.40-42 Tan Tsu Wee, T (2001) The use of Marketing Research and Intelligence in Strategic Planning: Key Issues and Future Trends. Marketing Intelligence and Planning 19(4) pp. 245-254 Van Scotter, J and Culligan P.E. (2003) The Value of Theoretical Research and Applied Research For The Hospitality Industry. Cornwll Hotel and Restaurant Administration Quarterly 44(2) pp.14 Read More