مدل های درخت تصمیم گیری برای پروفایل تبلیغاتی و استراتژی های تبلیغات اسکی و تاثیر در فروش
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|24111||2013||8 صفحه PDF||سفارش دهید||5620 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Expert Systems with Applications, Volume 40, Issue 15, 1 November 2013, Pages 5822–5829
Based on survey data, this paper builds decision tree models to profile the online and mobile technologies and services that ski resorts use for their promotional and advertising strategies for two important segments, namely millennials (less than and equal to 35) and non-millennials (greater than 35). The technologies and services include resort websites, microblogging services, and online coupon services. The decision tree models reveal that ski resorts use specific strategies for these segments. Also, the paper reveals the impact that the technologies and services have on resort sales. The impact is positive and both immediate and sustained in nature. The research is the first of its type in the ski industry and represents a novel use of decision tree models for profiling promotional and advertising strategies.
Ski resorts often use a variety of communication approaches, including advertising, sales promotion, and public relations, to communicate with their primary market segments. e-Commerce, m-commerce, and the proliferation of digital devices are making new digital communication channels available to ski resorts. For example, ski resorts can use resort websites, Groupon (i.e., an online coupon service), and Foursquare (i.e., a location-based service) for promoting and advertising to online followers and prospective customers. Currently, ski resorts are struggling to develop integrated promotional and advertising strategies that employ the available online and mobile technologies and services. This is due to the fact that that ski resorts do not understand current practices (i.e., what other ski resorts are doing) and how those practices affect sales. Knowledge discovery in databases (KDD) and data mining (DM), which combine databases and machine learning, may identify best practices (Fayyad & Stolorz, 1997). The primary KDD/DM steps are dataset construction, data mining (i.e., model building), model assessment, and interpretation of results. Other researchers (e.g., Wu, Kao, Su, & Wu, 2005) have used the methodology to analyze customer datasets for the purpose of generating decision rules for cross-selling or up-selling insurance products. Companies can acquire datasets from a variety of sources, including customer databases, internal transaction data (e.g., point of sale data), and/or survey data. Depending on their objective, including clustering, classification, or association, companies can apply one or more DM algorithms, including clustering, decision tree, and association rule algorithms, to build specific models to accomplish their objective. For example, if an objective is to identify customer segments, companies could use a customer dataset (e.g., customer account data) and the k-Means algorithm to develop a clustering model that would explicate customer segments. Several researchers (e.g., Levin & Zahavi, 2001) have used RFM attributes, for regency (i.e., time since last purchase), frequency (i.e. frequency of purchase), and monetary (i.e., total customer spend), respectively, to build and evaluate decision tree models for profiling and predictive modeling. The models identify the attributes (i.e., profiles) of customers that constitute predefined segments (e.g., buyers and non-buyers) and, in so doing, predict customer responses (e.g., buy). Although there are other customer-related attributes (e.g., age), RFM data is readily available in marketing and sales databases. As this previous research demonstrates, decision tree models offer a convenient and powerful way for profiling customers and building customer response models. This study uses decision tree models to profile the online and mobile technologies and services – hereafter referred to as just technologies and services – ski resorts employ for promoting and advertising to two customer segments, namely millennials, or Generation Y (less than or equal to 35 segment), and non-millennials (greater than 35 segment).1 The profiles reveal the promotional and advertising strategies that ski resorts use to communicate with two important demographic segments. Because this investigation is technology oriented, we exclude non-digital, communication approaches (e.g., direct mail campaigns). Thus, the research is about profiling promotional and advertising strategies rather than customers, as past research has done. To our knowledge, there are no empirical studies about the technologies and services that ski resorts use for promotional and advertising purposes, making this study a unique application of decision tree models. The study addresses several important research questions. First, what technologies and services do ski resorts use as part of their promotional and advertising strategies? Generally, anecdotal information is available in the popular press, trade magazines, and marketing and advertising blogs. Second, do the technologies and services employed by ski resorts have a positive impact on sales? Although many of the technologies and services may have a favorable impact on brand development, do they actually increase sales? Is the impact temporary or sustained in nature? After a review of related research, this paper overviews the technologies and services that ski resorts may use as part of their promotional and advertising strategies. Next the paper describes the dataset, decision tree algorithm, namely C5.0, and the actual decision tree models that result from applying the algorithm to the dataset. Then the paper discusses the impact of various technologies and services on ski resort sales. Finally, the paper discusses the implications for ski resorts, provides a summary, and offers several concluding comments.
نتیجه گیری انگلیسی
This research discusses the technologies and services that ski resorts may consider in promoting and advertising themselves. Based on survey data, the research reveals two decision tree models that profile the technologies and services that resorts use for their promotional and advertising strategies. The decision tree models reveal that ski resorts employ specific communication strategies for millennial and non-millennial segments. For example, for advertising purposes, ski resorts rely on social media and their websites to advertise to millennials. They augment these technologies and services with microblogging and text messaging to advertise to non-millennials. Our research also reveals that the communication strategies profiled by the decision tree models have had a favorable impact on sales. The impact may be immediate and long-term in nature. These findings suggest that ski resorts should include, in their promotional and advertising budgets, some – if not all – of the technologies and services discussed here, if they are not already doing so. At the very least, they should consider our findings as a benchmark for their constantly evolving promotional and advertising strategies.