We have checked the possibility of predicting the shopping needs of Poles using the analysis of the inhabitants of the Krzyki district in Wrocław as an example. The research results carried out jointly with MANDS - Market and Opinion Researches were presented at the 19th edition of the Congress of Market and Opinion Researchers in Warsaw.
The paths covered by consumers as well as the distances that separate the average clients from the stationery stores of one of three product categories (grocery store, DIY store, and drugstore) were placed under the precise analysis.
The main goal of the project was to determine which stores the Polish lack in their neighborhood. Thanks to the data analysis that shows the movement of residents and smartphone users' profiles in a given area, it is more likely to determine which store brand will be the best investment in the specified area. Data from mobile devices were collected in the period January-August 2018 based on 99 939 devices.
The Big Data analysis was compared with the results of a declarative survey carried out using the MAPI (Mobile Assisted Personal Interview) method implemented by the MANDS market research and opinion agency. Interviews with 150 residents of Krzyki were carried out via mobile devices.
During the project, based on the store maps where the residents of Krzyki make the most purchases, the average distances they cover to reach these establishments, and on the basis of behavioral profiling (comparison of 15 defined profiles between residents of the district and customers of the brands covered by the research), the brands were identified, the presence of which in the neighborhood is the more desirable.
The best-suited stores were those in which customers went shopping with as many common features as possible in a specific district. Among the profiles, are parents of children aged 1-4, women expecting a child, but also the Ekstraklasa league fans. The juxtaposition was compared with the results of a declarative survey, in which the residents were asked directly about what store they would like to have in their area.
The results of the analysis conducted for the Krzyki district showed that Big Data can support location decisions in the case of the network development. We can see the DIY stores as a prove. In this case, the average distance covered by customers to the facilities of the most popular network is 8062 meters, which is almost four times more than an equally well-known competitor in this industry. Nevertheless, the residents of the Krzyki district often visit the first-brand stores, to which they have longer distances (14% and 11% respectively).
“Analyzing the above results, we can recommend opening a new store of the more popular network in the district that has been analyzed. This is proved by declarative research” - says Marcin Augustyniak, Client Director of Data Tank, Selectivv.
"Big Data is large and diverse data sets. They show what customers do every day, where they shop, etc. However, we still have to answer the question - why do they behave like that? Here, for example, traditional interviews come into help, in which the respondent can always be asked why he thinks so. For example, such answers as "the wife decided", "I buy cat food there", "because it is next to the bus stop" appeared in the discussed project among the motivators of the store visits. Taking this into account, combining data from different sources is a good practice and we should use it as often as possible", - adds Przemysław Huk, Senior Research Executive, MANDS.
The study showed that the behavioral data obtained from the mobile devices provide a strong reason for building consumer predictions. The combination of information from the real world allows you to determine whether the offer of a shopping center, gallery, cafe or shop fits the profiles of people staying in the area and suits their needs. Of course, one should not forget that the best conclusions should be based on gained knowledge from many sources. In addition to data related to movement, one should remember about traditional research and about own experiences and intuition.
