تحلیل رفتار مشتری های خرده فروشی ها با استفاده از مدل های تعامل مکانی مطالعه موردی: شیرینی فروشی های شهر تهران

انتخاب مکان به‌منظور احداث یک فروشگاه جدید برای خرده‌ فروشی تصمیمی بسیار مهم است زیرا هزینه‌های زیادی را دربر دارد و فردی که فروشگاه جدیدی را احداث می‌کند، خود را در معرض خطر مالی قرار می‌دهد. موقعیت مکانی به خریدکردن اولیه‌ی مصرف‌کننده از یک فروشگاه و وفاداری نسبت به آن تاثیر می‌گذارد. از این‌رو تجزیه و تحلیل موقعیت مکانی برای فروشگاه‌های خرده‌فروشی بسیار اهمیت دارد. با این‌که انتخاب مکان برای یک خرده‌فروشی همیشه دشوار بوده است، وضعیت رقابتی کنونی هم این تصمیم‌گیری را دشوارتر کرده است، زیرا فروشگاه‌ها به‌طور گسترده‌ای، با رقابت زیاد مواجه هستند. بنابراین تصمیم‌گیری برای یافتن محل یک فروشگاه جدید نیازمند یک راهبرد مکانی است. بسترهای خدمات‌دهی برخط تابع یک سری قیود هستند: به‌عنوان مثال، فقط به بخشی از شهر خدمات ارائه می‌کنند و این امر باعث می‌شود مدل‌های تعامل مکانی را نتوان بر روی کل شهر اجرا کرد. لذا در این مقاله با تکیه بر مدل تعامل رقابتی ضربی از نظریه مکان خرده‌فروشی، یک مدل بازاریابی مبتنی بر مکان برای خرده‌فروشی‌ها توسعه داده شده است که به راهبردهای مکانیابی برای احداث یک شیرینی‌فروشی جدید کمک می‌کند. ابتدا ویژگی‌هایی که در جذب مصرف‌کننده به شیرینی‌فروشی‌ها تاثیر دارد، تعیین می‌شود. سپس با استفاده از ابزارهای سیستم اطلاعات مکانی داده‌ها مورد تجزیه و تحلیل قرار گرفته و مدل پیاده‌سازی می‌گردد. نتایج تحقیق نشان می‌دهد که مدل پیشنهادی توانسته است با میانگین خطای 17.03 درصد به پیش‌بینی رفتار مصرف‌کننده بپردازد و به افراد در مکانیابی فروشگاه جدید با توجه به ویژگی‌های فروشگاه، رقبا و محیط کمک کند. مدل پیشنهادی این تحقیق می‌تواند برای افزایش دقت در مکانیابی مراکز خرید دیگر هم استفاده شود.

Analyzing the behavior of retail customers using spatial interaction modelsCase Study: Confectionaries in Tehran

Extended AbstractIntroductionSelecting a suitable place for a new retail store is a very important decision since new shops cost a lot and new retailers puts themselves at financial risk. Physical location of stores affects the consumer’s perception of their first purchase and their subsequent loyalty to the store. Therefore, spatial analysis is very important for retail stores. Site selection for retail stores has always been difficult and the current competitive market has made decision making even more difficult since stores face increased competition and consumers have many options to satisfy their needs. They generally choose a suitable store in their vicinity which provides high quality, cheap, and diverse products. Therefore, markets and especially retailers shall follow an accurate and valid location strategy for new stores. Retail stores have various marketing and customer service strategies. Marketing strategies require a lot of information about different aspects such as customers, shops, competitors, and products. Many marketing strategies only provide information about consumer behavior or customer satisfaction. However, spatial aspects are more important and in fact determine future success of a store. Several methods are used for spatial analysis in retail sector. The present study use a multiplicative interaction model to forecast sales of confectionaries. This can help retailers develop strategies and find an optimal location for their new stores. Materials & MethodsThe present study has developed a locationbased marketing model for online confectioneries in Tehran which can improve site selection strategies of new confectioneries. This marketing model is based on the multiplicative competitive interaction model (MCI) of the retail location theory. To do so, characteristics attracting customers to confectioneries are determined and related data are collected from the Snappfood online platform through web crawling. ArcMap software is then used to analyze and process the collected data. After data normalization, MCI model is implemented using Python programming language. The model is then calibrated using 80% of the collected data and the ordinary least squares (OLS) method. The model is then evaluated using root mean square error (RMSE) method and the remaining data. Results and DiscussionMean errors obtained for districts number 1 to 22 of Tehran municipality show high accuracy of the model. Snappfood site lacked any information about districts number 9 and 18 and thus these districts were not considered in the calculations. Depending on the available data, other districts showed different levels of accuracy. Results indicate that district number 22 had the lowest level of accuracy and district 17 had the highest level of accuracy.  In general, this model predicts customer behavior with an error rate of 17.03%. Results of the present study show the probability of purchasing from each confectionery which can be used to map market potential for a new store. This map determines the best place with maximum sale and helps in site selection for new stores based on specific features of the store, competitors and the environment. ConclusionsMCI model predicts sales. From a geomarketing perspective, this model shows that distance between customers and the store and accessibility affect location strategies in new stores. Variables such as pricing and customer satisfaction (scoring) are used to improve the goodnessof fit of the model. This precise method identifies some key factors to success in a retail strategy. It predicts the probability of purchasing in each district, the number of customers in each store, and distribution of customers in each district. Experts and new retailers can use the results to design various location and sales strategies. Using this model, new retailers in confectionary market can accurately predict their sales before even opening the store and thus protect themselves against possible financial losses. Moreover, this model predicts total sales of different stores and help retailers compare their market shares with those of their competitors. They also can enter features of a new store into the model and find several potential sales strategies. In other words, the model helps determine sales of existing and new shops. In this way, retailers can find an optimum location for their new confectioneries based on the principles of geomarketing.