Data-driven urban commerce management: The case of Villena

GRADO: Administración y Dirección de Empresas Curso 2021/2022 DATA-DRIVEN URBAN COMMERCE MANAGEMENT: THE CASE OF VILLENA Autor: Julen Pérez Blanco Director: Jon Charterina Abando Bilbao, a 24 de junio de 2022 Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 2 Table of contents 1. INTRODUCTION ......................................................................................................... 8 1.1. PURPOSE AND OBJECTIVES OF THE INVESTIGATION ................................................... 8 1.2. SCOPE .......................................................................................................................... 9 1.3. JUSTIFICATION OF THE PROJECT ................................................................................. 9 2. REVIEW OF THE LITERATURE .................................................................................... 10 2.1. URBAN COMMERCE REVITALIZATION PROGRAMS ................................................... 10 2.2. URBAN COMMERCIAL CENTRE .................................................................................. 13 2.3. DATA-DRIVEN MANAGEMENT .................................................................................. 15 3. CONTEXTUALIZATION OF THE CASE .......................................................................... 15 3.1. VILLENA ..................................................................................................................... 15 3.2. 03400 VILLENA APP ................................................................................................... 18 4. METHODOLOGY ....................................................................................................... 19 4.1. TIMELINE OF THE INVESTIGATION ............................................................................. 20 4.2. SOURCES OF INFORMATION USED ............................................................................ 21 4.3. QUANTITATIVE TECHNIQUE: SURVEY ........................................................................ 21 4.3.1. Information collection system .......................................................................... 21 4.3.2. Description of the survey ................................................................................. 22 4.3.3. Description of the sample obtained from the survey ....................................... 23 5. RESULTS ................................................................................................................... 24 5.1. UNIVARIATE ANALYSIS .............................................................................................. 25 5.2. BIVARIATE ANALYSIS ................................................................................................. 32 5.2.1. Description of the youngest demographical group .......................................... 32 5.3. MULTIVARIATE ANALYSIS .......................................................................................... 33 5.3.1. Characterisation of consumers in Villena by profiling ..................................... 33 5.4. ANALYSIS OF OPEN-ENDED QUESTIONS .................................................................... 36 6. CONCLUSIONS, DISCUSSION, RECOMMENDATIONS AND LIMITATIONS .................... 38 6.1. CONCLUSION ............................................................................................................. 38 6.2. DISCUSSION AND RECOMMENDATIONS ................................................................... 39 6.3. LIMITATIONS AND FUTURE LINES OF INVESTIGATION .............................................. 41 BIBLIOGRAPHY .................................................................................................................. 42 ANNEX .............................................................................................................................. 45 ANNEX 1: POPULATION DENSITY AND AVERAGE AGE OF POPULATION IN 2019 CONTRASTED WITH MAIN AFFLUENCE AREAS .................................................................................................................... 45 ANNEX 2: QUESTIONNAIRE ........................................................................................................ 46 ANNEX 3: DIFFERENTIATING VARIABLES FOR CLASSIFICATION BY CLUSTER ANALYSIS ................................ 51 ANNEX 4: CROSS TABLES BETWEEN CLUSTERS AND SOCIO-DEMOGRAPHIC VARIABLES ............................. 54 ANNEX 5: CROSS TABLES OF VARIABLES USED FOR CLUSTER CHARACTERIZATION .................................... 56 Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 3 List of tables: Table 1: Advantages and disadvantages of urban commerce depending on location ........... 11 Table 2: Timeline of the investigation ..................................................................................... 20 Table 3: Descriptive statistics of variables related to customer satisfaction .......................... 26 Table 4: Differentiating variables for classification by cluster analysis ................................... 34 Table 5: Summary of the comparison in each cluster ............................................................. 34 Table 6: Characterization of clusters with sociodemographic variables ................................. 35 Table 7: Characterization of clusters with more variables ...................................................... 35 Table 8: Categorization of the open-ended questions ........................................................... 36 Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 4 List of figures: Figure 1: Annual rate of change in household final consumption expenditure ...................... 11 Figure 2: Variables affecting urban management ................................................................... 14 Figure 3: 2021 aerial image of Villena ..................................................................................... 16 Figure 4: 2021 population pyramid of Villena ........................................................................ 16 Figure 5: Logo of Asociacion de Comerciantes de Villena ....................................................... 17 Figure 6: Percentage of the sample according to gender ....................................................... 23 Figure 7: Percentage of the sample according to age ............................................................. 23 Figure 8: Percentage of the sample according to familiar situation ....................................... 24 Figure 9: Percentage of the sample according to place of residence ..................................... 24 Figure 10: Frequency distribution to the question: “How often do you shop at Villena urban commerce?” ........................................................................................................................... 25 Figure 11: Frequency distribution to the question: “What is the main reason for making these purchases?” ................................................................................................................... 25 Figure 12: Frequency distribution to the question: “What type of purchase do you usually make?” .................................................................................................................................... 26 Figure 13: Frequency distribution to the question: “How often do you shop at the municipal marketplace?” ......................................................................................................................... 27 Figure 14: Frequency distribution to the question: “What type of purchase do you usually make at the municipal marketplace?” .................................................................................... 28 Figure 15: Frequency distribution to the question: “How often do you shop at Thursday’s Street market?” ...................................................................................................................... 28 Figure 16: Frequency distribution to the question: “What type of purchase do you usually make at Thursday’s Street market?” ...................................................................................... 29 Figure 17: Frequency distribution to the question: “Which shopping mall do you usually visit?” ...................................................................................................................................... 29 Figure 18: Frequency distribution to the question: “What type of purchase do you usually make online?” ......................................................................................................................... 30 Figure 19: Frequency distribution to the question: “What type of purchase do you usually make at shopping malls?” ....................................................................................................... 30 Figure 20: Frequency distribution to the statement: “I buy in the local commerce of Villena because:” ................................................................................................................................ 31 Figure 21: Frequency distribution to the statement: “The main problems of commerce in Villena are:” ............................................................................................................................ 31 Figure 22: Type of purchase do you usually make do young people make at Villena urban commerce ............................................................................................................................... 33 Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 5 Abstract Local commerce has been impacted by various external factors over the last decades. From the changeover to the euro in the early 2000s, to the great economic recession in 2008, to a global pandemic in the last two years. During this period, society has continually evolved, changing tastes and preferences, which has been a major challenge for small businesses. In this way, local trade has been studied by economists and institutions in order to push it towards new trends. The presence on the Internet is increasing, which is why the digital transformation is becoming so important. New technologies provide a wide range of opportunities for commerce, but they must be used correctly, otherwise they will be incurring on a cost. In this context, data and data-driven management is one of the main pillars by which local councils, associations and businesses are being guided. The research has been carried out in several stages. Firstly, a contextualization of urban commerce and open shopping centers was carried out. In addition to explaining the many concepts related to the subject, a presentation of the city under study was made. Finally, a qualitative research was carried out in collaboration with the marketing consulting company Neuromobile and Asociación de Comerciantes de Villena. Thanks to the cession of the data collected through a survey carried out by them, it has been possible to accomplish an analysis of the same, obtaining several conclusions that will be presented at the end of the study. Keywords: local commerce, smart city, data, people, shopping experience, Villena. Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 6 Resumen El comercio local se ha visto afectado por diversos factores externos en las últimas décadas. Desde el cambio al euro a principios de la década de los 2000, pasando por la gran recesión económica de 2008, hasta una pandemia mundial en los últimos dos años. Durante este periodo, la sociedad ha evolucionado continuamente, cambiando gustos y preferencias, lo que ha supuesto un gran reto para el pequeño comercio. Así, el comercio local ha sido estudiado por economistas e instituciones para impulsarlo hacia nuevas tendencias. La presencia en Internet es cada vez mayor, por lo que la transformación digital está adquiriendo una gran importancia. Las nuevas tecnologías aportan un amplio abanico de oportunidades al comercio pero se debe hacer un correcto uso de estas, ya que si así no fuera, se estaría incurriendo en un coste. En este contexto, el dato y la gestión a través de este es uno de los principales pilares por el cual tanto ayuntamientos como asociaciones y comercios se están guiando. La investigación se ha llevado a cabo en varias etapas. En primer lugar, se ha realizado una contextualización del comercio urbano y de los centros comerciales abiertos. Además de explicar los numerosos conceptos relacionados con el tema, se realizará una presentación de la ciudad objeto de estudio. Por último, se llevará a cabo una investigación cualitativa en colaboración con la empresa de consultoría de marketing Neuromobile y la Asociación de Comerciantes de Villena. Gracias a la cesión de los datos recogidos a través de una encuesta realizada por ellos, se ha podido realizar un análisis de estos, obteniendo varias conclusiones que se presentarán al final de la tesis. Palabras clave: comercio local, ciudad inteligente, datos, personas, experiencia de compra, Villena. Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 7 Laburpena Lekuko merkataritzari hainbat kanpo-faktorek eragin diote urteotan. 200ko hamarkadaren hasieran eurora aldatu zenetik, 2008ko atzeraldi ekonomiko handitik, azken bi urteotako mundu mailako pandemiaraino. Aldi horretan, gizarteak etengabe eboluzionatu du, bere gustuak eta lehentasunak aldatuz, eta hori erronka handia izan da saltoki txikientzat. Horrela, ekonomialariek eta erakundeek tokiko merkataritza aztertu dute joera berrietara bultzatzeko helburuarekin. Interneteko presentzia gero eta handiagoa da, eta, beraz, eraldaketa digitala garrantzi handia hartzen ari da. Teknologia berriek aukera ugari ematen dizkiote merkataritzari, baina behar bezala erabili behar dira, horrela ez balitz, kostua izango litzateke. Testuinguru horretan, udalak, elkarteak eta dendak datua erabiliz eta kudeatuz gidatu behar dira. Ikerketa hainbat etapatan banatu da. Lehenik eta behin, hiri-merkataritza eta merkataritza-gune irekiak testuinguruan kokatuko dira. Gaiarekin lotutako kontzeptu ugariak azaltzeaz gain, aztergai den herriaren aurkezpena egin da. Azkenik, ikerketa kualitatibo bat egin da Neuromobile, marketin-aholkularitzako enpresarekin eta Asociación de Comerciantes de Villena-rekin lankidetzan. Haiek egindako inkesta baten bidez jasotako datuen lagapenari esker, datuen azterketa egin ahal izan da, eta tesiaren amaieran aurkeztuko diren hainbat ondorio atera dira. Gako-hitzak: tokiko merkataritza, hiri adimenduna, datuak, pertsonak, erosketaesperientzia, Villena. Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 8 1. INTRODUCTION Society is constantly evolving, and with it, new trends and preferences to satisfy. Local commerce is an economic activity that has had to analyse these emerging trends to adapt its offer to customers' needs. Additionally, they have come up against a very difficult competitor, the Internet and its online shopping platforms. Shopping portals such as Amazon or online shops have made customers prefer to wait a few days to receive their products rather than pay a higher price in their local shops. In addition, the current situation, marked by the Covid-19 pandemic, has pushed the vast majority of shops to their limits, and in many cases, they have had to close. Businesses have had to join forces and make a maximum effort to find solutions to the loss of customers. For the reason that local commerce has to compete against the world’s largest companies, they need to rely on business associations, institutional support and professional advice. Setting programs up as quickly as possible might sound a proper solution. Even if the programs are different from each other, all of them should aim to solve the existing problem. However, it is essential to understand the reality in as much detail as possible to be able to design those plans and data analysis is the first step for that. 1.1. PURPOSE AND OBJECTIVES OF THE INVESTIGATION In the following lines, the purpose of the present work will be presented in the first place. Next, the objectives, which are divided into general (objectives) and specific (subobjectives). We will try to answer the proposed objectives through the quantitative part of the investigation, by analysing the results of the survey. The objective of this study is to analyse the urban shopping area in a medium-sized town. The aim is to discover the profile of the person who purchases locally and for this purpose we are going to study the municipality of Villena in Alicante, Spain. To do this, a comparison will be made of how customers behave in different places where they can make their shopping. In addition to this, we are also going to find out how factors such as proximity, price or trust exert an influence on customers when making their purchases. Moreover, this study aims to offer solutions to the problems that customers have expressed in the survey, to help the commerce of the town improve customers’ shopping experience and consequently, the number of sales. It is also intended to try to analyse the factors why local shops are losing customers in the younger age groups, as a first step to designing strategies to retain and attract this young group to participate in the local economy. • Objective 1: describing the behaviour of individuals in the local commerce of Villena, as well as their tastes and preferences. o Sub-objective 1.1: defining the frequency of purchases on the urban commerce of Villena. o Sub-objective 1.2: describing the product preferences on the urban commerce of Villena. Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 9 o Sub-objective 1.3: describing how the younger demographic group (18-44 years old) interacts with local commerce in Villena. o Sub-objective 1.4: to segment Villena's shoppers into divergent groups and to describe each of those groups. • Objective 2: to ascertain which are the main purchase motivators in the local commerce of Villena. o Sub-objective 2.1: Finding out if the factors that act as purchasing levers vary depending on the age of the person. • Objective 3: measuring the satisfaction level of the clients with the local commerce of Villena. o Sub-objective 3.1: understanding the problems that local commerce of Villena faces. o Sub-objective 3.2: getting to know personal opinions about what type of products people miss on Villena’s local shops. • Objective 4: design a series of recommendations to the town council and Asociación de Comerciantes de Villena to improve their urban shopping experience and the services offered. The objectives pursued in this research have been designed by the author together with the collaborating company and the association of commerce in Villena, under the supervision of the professor tutoring it. These have been modified throughout the process to meet the needs of the stakeholders. 1.2. SCOPE To develop this research, information has been sought in various manuals, magazines, websites, among others, to develop the contextualization of urban commerce and the literature review. In addition to this, we have also used the data obtained from a survey carried out in January 2022 in the town of Villena and which have been provided by the collaborating company in this work, Neuromobile. In addition, both Asociación de Comerciantes Villena and the town council have provided reports and written documents that have helped in the preparation of this document. As mentioned above, the study has been carried out on the data obtained from a survey conducted by Neuromobile in the town of Villena in February 2022. Although the vast majority of the participants are residents of the town, it is worth mentioning that there is a slight participation of both tourists and people from neighbouring towns, people living in the province and people who spend their holidays in the village. 1.3. JUSTIFICATION OF THE PROJECT The continuous changes that local commerce has undergone in recent decades, together with the uncertainty brought by COVID-19, have led many local councils to start Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 16 Figure 3: 2021 aerial image of Villena Source: Wikimedia Commons (2008) Villena has a network of roads that connects it with nearby town and cities, which also includes a highway that connects directly with Alicante on side and with the capital of Spain, Madrid on the other one. It also has a commuter train station, and a high-speed station that is connected to the Spanish high-speed train network. With a land surface of 345,37 km2 there was living 34.025 people in the city back in 2021. Population density varies depending on the commercial zones into which the town is divided. As can be seen in Annex 1, Villena is divided into eight main areas of affluence, and it is the most central ones that have the highest population density. Thus, affluence areas 1, 2, 4, 5 and 6 are where most people live. The average age of residents in the town also varies depending on the commercial zones. The second picture on the same annex shows how older people tend to live in the central shopping areas, while younger people tend to live in shopping areas further out, such as 3 or 7. There is a reason for this, which is that housing is cheaper on the outskirts than in the historic centre of the town. It has relevance on the region as 64,74% of the habitants live there. The population pyramid below shows the population by five-years groups and gender in Villena on 1st January 2021. (Ayuntamiento de Villena, n.d.) Figure 4: 2021 population pyramid of Villena Source: compiled by author based on INE The figure displays a pyramid with a regressive shape, which means that the population tends to be middle-advance aged. This happens because there is a large group of people living -10,00% -5,00% 0,00% 5,00% 10,00% 0-4 15-19 30-34 45-49 60-64 75-79 90-94 Female Male Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 17 in Villena, which are between 30 and 64 years old, representing the 50,4% of the people living in the town. The group from 15 to 29 is the 14,94% and the ones over 64 years old is 18,62% (Instituto Nacional de Estadística, 2021). Regarding economic activity, there were 2.243 companies in Villena in 2021, from which 11,77% were dedicated to industrial activities and 12,8% to construction. The main economic activity in the city are services, which represent 42,44% of the companies in the city. (Empresas por municipio y actividad principal(4721), n.d.). The unemployment rate on 31st January 2022 was 13,13%, almost 3% superior than the average on the Valencian Community (Datos Estadísticos | Trabajo | Villena - ARGOS, n.d.). The city has registered 35.384 tourists during 2021 which is a larger number of tourists than before the global pandemic. This has a moderate impact on the economy of Villena, which has been affected during the last couple of years (Prats, 2022). The commercial amount indicator shows there are 19 establishments per 1000 habitants. 251 are classified as traditional alimentation commerce, 120 are focused on apparel and footwear, 96 to homeware. Additionally there are 9 supermarkets, 1 hypermarket and 22 mixed classified commerce (Ayuntamiento de Villena, n.d.). They also have a commercial association named Asociacion de Comerciantes de Villena (from now on, it will be referred as ACV). It stands for conducting the town into a shopping city, offering the citizens the opportunity of having some fun while purchasing at their urban shopping centre. To achieve that, the association defends and represents the business interests, improves capabilities and decision making, communicates by continuous diffusion, and saves costs; always driven by the principles of cooperation, participation, environmental compromise, equality, and innovation. (Asociación de Comerciantes de Villena, n.d.). Figure 5: Logo of Asociacion de Comerciantes de Villena Source: Asociación de Comerciantes de Villena 2 This Association, in collaboration with the Villena Town Council, launched the open space shopping centre project in 2005. It was born with the purpose of improving the accessibility, urban planning, lighting, cleaning and parking of the commercial streets of the city. This open space shopping centre has been so successful that in 2017 the Ministry of Economy and Competitiveness awarded it the National Interior Commerce Award in the category of Open Shopping Centres. This award rewards the promotion and modernization of the sector as well as the partnership shown by all parties involved in the project (Alcaraz, 2017). 2 The image has been provided by Asociación de Comerciantes de Villena for use in this study. <15/03/2022> Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 18 3.2. 03400 VILLENA APP As it is stated on Neuromobile`s web page, 03400 Villena is a digital platform that’s has been launched as an app which has become into a solution of communication with citizens that provides services to different areas of the town hall. From festivals to tourism or economic development, this platform emphasizes the relevant role that local commerce has within the economic and cultural activity of the city and the importance of improving urban commercial environment management through data. This solution was implemented by Neuromobile back in May 2020 in the middle of Covid-19 pandemic to boost the local economy by enabling new channels of communication with citizens. The outbreak had a relevant role on the initial success of the app because it provided the information of the evolution of the pandemic in the city, which was part of the new services that were essential during that period. When the app was launched, 2000 initial downloads were achieved. As Asociación de Comerciantes de Villena indicates, women are the ones that use most this app, since only 38,42% of the users are men. Moreover, 43% of the users are older than 45 years, which shows that this segment has find the app useful. There are more than 350 businesses participating in the app, which has also helped growing ACV in 50 more members. As this platform of capturing data enables a large flexibility to offer services to citizens, together with the town hall and with ACV they have added new services to capture new segments. By adding local restaurants home delivery services and management of tickets for summer concerts and festivals they have pushed young people from 18 to 25 people into local shopping. This app is an implementation of the Dynamic Urban Commerce (DUC) solution offered by Neuromobile. This company was born between 2013 and 2014, through an idea that raised during Start-up Weekend contest in Murcia. It was first designed to serve as platform to develop marketing actions and dynamics for shopping malls. However, in 2018 in turned into a management platform that helped on decision making. They were totally settled on the shopping mall sector, so decided to adapt the capabilities of the platform to help urban areas getting dynamized. Large research was made across Spain, trying to find associations that were trying to enliven and activate local commerce. It was this moment when they contacted ACV, and the DUC solution was introduced to the association. Since the objectives of the association met with the vision of the company, they began to cooperate and 03400 Villena App platform was released as an adaptation of the DUC solution. DUC is a configurable, unified, and smart platform that provides the tools to manage the territory, linking the city, with its businesses, residents, and occasional tourist. It is designed to improve the shopping experience and boost the participation of the population of the geographical area the tool is targeted on. The dynamization of commercial areas is achieved by obtaining market data and analysing it to understand customer behaviour in the commercial space. By studying the collected data, dynamic marketing campaigns are created, and the tools built inside the platform help retailers to make decisions. The purpose of this platform is to become the solution by integrating a powerful set of digital tools (BIGDATA) with a simple and intuitive interface that serves as a platform of administration and performs data management (SMARTDATA) to assess and measure the impact of the activities in which the resident and occasional citizen participates. Due to its Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 19 architecture and development, it is a project that can be implemented rapidly and is easily scalable. DUC also serves as a simple communication interface with the consumer. It is designed to be the only contact point between the client and the manager where information about local economic and cultural activities is shared. Through the platform, the customer also is provided with exclusive content and offers that are not available at any other medium. Furthermore, bidirectional personalized communication between local businesses and administrations is offered to the inhabitant. It means that through the platform the client can directly contact businesses or administration. This leads into facilitating consumers to participate in surveys or inquiries and improving incident management and communication. As mentioned, DUC provides technicians and managers of the urban environment with multiple levers of action such as bonds, coupons or discounts that promote economic activity. Example of those levers are segmentation of content from a single repository or facilitating connection with digital channels such as websites or social networks. Those levers also help to create dynamization plans, execute of them and monitor the objectives and main key performance indicators (KPI) of the results obtained. This solution follows the principles of the Sustainable Development Goals since it advocates for a sustainable urban and rural development through an intelligent tourism 3 . Thanks to this application, all activities related to encouraging shopping in Villena's local shops, such as promotions and prize draws, have been automated. This has made it possible to reduce the association's promotional costs by 35%. It also serves as a data collection platform. As the application registers users, it also registers their behaviour with local commerce. For example, it is possible to obtain the average ticket of the participants in the different campaigns that are carried out. The “Villenear” campaign totalled 15,066 tickets with a value of 604,051€. By recording the data, it can be seen that the average ticket was 40€ and that the person with the most tickets has registered 218 different participations or that the ticket with the lowest value was 0.10€. As data management is one of the objectives of both Neuromobile and Asociación de Comerciantes de Villena, a survey was launched through this platform in January 2022. The aim of the survey was to understand the shopping preferences of people in the town in order to offer them the best possible service. Once the survey was over and the information had been compiled into a database, it was time to analyse the data in order to draw conclusions for improvements. 4. METHODOLOGY In this section of the work, the different techniques used to carry out the research about the concrete and real image of urban commerce in Villena will be explained. Firstly, the process followed from start to finish will be presented, so that a clear vision of the workflow is provided. Secondly, the sources of information used will be mentioned, and finally, the survey will be alluded. 3 Information obtained from various documents provided by the company Neuromobile. <10/02/2022> Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 20 4.1. TIMELINE OF THE INVESTIGATION Once the case was introduced to the author of the study, and was accepted by him during mid-February, it was decisive to design a timeline. Thus, there is a greater orientation towards responding the proposed objectives and drawing the appropriate conclusions and recommendations. For each phase of the investigation a guiding timeline has been set, which has been modified depending on time. It is also worth mentioning in this section the involvement of the Neuromobile team, as well as town council workers and members of Asociación de Comerciantes de Villena. Several meetings have been held with them, which have served to guide the work. In the first meeting, both students and teachers met the people involved in the task and were informed of the objectives of the project. In the following meetings, the progress made in the research was reported, as well as making small adjustments as they arose during the development of the project. Table 2: Timeline of the investigation Tasks Jan-22 Feb-22 Mar-22 Apr-22 May-22 Jun-22 Jul-22 Information search Case introduction Contextualization Objectives Univariate analysis Bivariate analysis Open-ended answers analysis Multivariate analysis Conclusion and recommendations Delivery Presentation Source: elaborated by author Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 21 4.2. SOURCES OF INFORMATION USED As mentioned in section 1.2 of the project, the information used to carry out this work has been consulted from various sources. Firstly, for the literature review part, manuals, journals, websites, integral plans of various Autonomous Communities and INE data have been used. All the sources consulted have been included in the bibliographical section, prior to the annex. Secondly, data obtained from the survey carried out in Villena has been used. This information was shared through an excel sheet provided by Neuromobile. At this point, I would like to thank the people involved in the project both for providing their work and their data. I would like to mention Marcos García and Raul García, both CEO of the company Neuromobile and María Ángeles García and María José Sauco, both part of the administration and management of Asociación de Comerciantes de Villena. In addition to providing the data, these people have also contributed with information about the town of Villena and various technical data about the survey. 4.3. QUANTITATIVE TECHNIQUE: SURVEY In this last part of the methodology, technical data on the survey performed in Villena will be presented. Consequently, it will be explained how the information collection process was carried out, into which parts the survey was divided and finally and the sample obtained will be described. 4.3.1. Information collection system The platform used to carry out the online questionnaire has been Tally, a tool to create forms recommended by Neuromobile. This program allows making attractive forms for the user by including images, videos or emojis. The results obtained are connected with the DUC platform for the automatic updating of customer data. The survey could only be solved through the app 03400 Villena and to encourage participation among citizens, the first one hundred individuals who answered were awarded. It was open for response between January 25th and February 1st, 2022. The questionnaire between ACV and Neuromobile, was disseminated through various channels. First, a press release was sent to local and regional media on January 27, 2022, which was published by the newspaper of Villena, Cadena Ser and Confecomerç among others. Two WhatsApp channels were also used, one to send the survey to customers and another to send it to establishments associated with ACV. The survey was also diffused through Facebook, Instagram, and Twitter. In addition, within the 03400 Villena app, push messages and a MOF in the form of a button were created to carry out the survey from the app. Finally, a newsletter was used to send the survey to app users and partners. Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 22 4.3.2. Description of the survey The questionnaire consists of six parts and was designed by the company Neuromobile to help the association know the customer and their assessment of local commerce. Moreover, their aim is to understand their buying habits in the face of the evident change in trends, visible especially after the pandemic. The survey is divided in the 7 different but connected parts. Part 1 is designed to obtain a sociodemographic description of the buyer in the urban commerce of Villena. It is made up of eight questions where name, surname, gender, age, familiar situation, place of residence, phone number and name are asked. Not all the information is relevant for this research as only gender, age, familiar situation and place of residence will be valuable for the description of the sample obtained. The objective is to know possible common characteristics among individuals in order to segment the target audience. The following part, Part 2, is composed of thirteen questions and tries to respond general aspects related to shopping at commerce in Villena. Respondents are asked about the frequency of their purchases, and which are the main reason for doing them. In the third part of the questionnaire, individuals are asked if they buy products from the following categories: groceries, beauty/health, fashion/complements, home/decoration and leisure. they give an affirmative answer, they must rate their level of satisfaction with the product category from 1 to 5. Part 4 is dedicated to the municipal marketplace of Villena. This place is a municipal building that currently has 25 food stalls open. Twelve questions have been formulated in order to gather data about shopping habits in this place. They are asked first whether they buy there or not and only will answer the questions regarding the municipal marketplace if they response affirmatively. Questions are related to expenditure, frequency and product categories. Part 5 is similar to the previous one, but surveyed people are asked about the street market held on Thursdays. They are asked first whether they buy there or not and only will answer the questions if they give an affirmative response. Then, eight questions related to frequency, product categories and expenditure appear. Part 6 is a fourteen-question brief investigation about online shopping habits and purchasing at shopping malls. The frequency of the purchases as well as the product categories have been analysed. Additionally, people who shops at malls are asked to clarify which shopping mall do they visit. The objective of the last part, part 7, is getting to know the perceptions and opinions of the individuals regarding their motivations and problems they encounter. With the aim of giving a response to those objectives, respondents are given some options to choose from regarding their reasons for purchase. Same method is used when asking about the main problems of the city of Villena. There also exist a final open question where individuals have the opportunity to write about what kind of products or brands do they miss in the local commerce of Villena. A copy of the survey used is attached to the annex 2. Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 23 4.3.3. Description of the sample obtained from the survey Sociodemographic variables (gender, age, family situation, place of residence) are key to understand the characteristics of the sample obtained. The results of the surveys have been collected in an Excel table from which a series of graphics have been made for a correct interpretation of them. Figure 6: Percentage of the sample according to gender Source: compiled by author based on obtained data The first variable that is going to be analysed is gender. As it is shown on the graphic, almost 59% of the people who completed the survey was female whereas only 13% was male. This means the number of women has been larger than men. However, there is a considerable number of participants that defined themselves as other. Figure 7: Percentage of the sample according to age Source: compiled by author based on obtained data The second demographical variable which has been analysed on the survey is the age of every participant. As it can be observed there is a greater participation among older people than younger people. This happens because 61% of the sample is concentrated between people from 35 to 54 years old. Another significant group is people between 55 and 64 years which represent 16,6% of the total. There is even a slight participation among people who is 65 years or older. The first age groups (18-24 and 25-34) have also participated in a 19,1%. Taking everything into consideration, it can be stated that the sample obtained from the survey is considerably distributed. 13% 59% 28% Male Female Other 6,60% 12,50% 33,60% 27,40% 16,60% 3,20% 0,00% 10,00% 20,00% 30,00% 40,00% 18-24 25-34 35-44 45-54 55-64 >=65 Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 24 Figure 8: Percentage of the sample according to familiar situation Source: compiled by author based on obtained data The next demographical variable is familiar situation. It has been divided into being single, being with a couple or in a family. People living in families account for just over half of the total. It is followed by couples representing 19% and finally people who is single picturing 27%. This distribution is coherent with the one made according to the age, as people from 35 to 64 tend to live on families. Figure 9: Percentage of the sample according to place of residence Source: compiled by author based on obtained data The last variable that has been analysed is the place of residence. Since this survey was designed to get to know the opinion of how to improve commerce in Villena and its environment, it is logical that the largest number of participants do live in the city (94%). There are also some participants who live in neighbouring villages (3%) and others that come to Villena to live during weekends and holidays (2%), but they are not relevant. 5. RESULTS During this stage of the investigation, the results obtained from the survey will be examined. Quantitative results will be analysed following univariate, bivariate and multivariate techniques. In the end, the open-ended question will be also analysed. 27% 19% 54% Single Couple Family 94% 2% 0% 3% 0% 1% Living in Villena Lives during weekends and holidays in Villena Lives in the province Lives in a neighboring village Tourist Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 25 The micro-data obtained through the questionnaire was collected in an Excel table by Neuromobile. They shared this document with 559 valid answers in February 2022 through a confidentiality agreement with the research report tutor, Jon Charterina. The professor dedicated himself to coding these answers to be able to analyse them using the IBM SPSS Statistics software. Once the SPSS results were obtained, they were sent to the author of this final degree project so that he could deeply analyse them. 5.1. UNIVARIATE ANALYSIS The first analysis is going to be the univariate, as it allows studying the variables of the survey individually. For some of them, the average, the median, the standard deviation and the maximum and the minimum of the data under study have been extracted. The following graphs refer to the second part of the survey, where to know the experiences of urban local shopping in Villena, individuals were asked about their habits in order to delimit and understand their perceptions and preferences (obj. 1). Figure 10: Frequency distribution to the question: “How often do you shop at Villena urban commerce?” Source: compiled by author based on obtained data Figure 11: Frequency distribution to the question: “What is the main reason for making these purchases?” Source: compiled by author based on obtained data These two graphics, analyse two of the questions raised about the shopping experience. In the first one, it can be observed that the great majority of the sample, 90,9%, shops at Villena urban commerce daily and weekly (sub-objective 1.1). 46,20% 44,70% 5,70% 3,40% 0,00% 10,00% 20,00% 30,00% 40,00% 50,00% Daily Weekly Monthly Ocasionally 54,90%26,80% 16,50% 1,80% Trust with shopkeepers Time saving Product quality Best price Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 32 5.2. BIVARIATE ANALYSIS The following analysis, the bivariate analysis, aims to find out the existing relationships between different relevant variables in the research compared to the sociodemographic variables that surveyed people have indicated. 5.2.1. Description of the youngest demographical group The first bivariate analysis is aimed at describing the demographic group considered to be the youngest. Therefore, all the questions proposed in the questionnaire were related by using cross-reference tables to the age of the surveyed. This demographical separation of the results helps into responding sub-objective 1.3. People from 18 to 44 years of age are considered to be part of the young demographical group. By contrast, people over 44 years old are the ones considered to be elderly. The vast majority of young people shop in local shops in Villena on a weekly (51.9%) and daily (36.6%) basis. This distribution is also found in the older demographic group, so it is not a difference that characterizes them. These results are due to the fact that, as will be seen below, the products purchased in the market are usually staple foods, which need to be purchased with a high frequency. However, young people are less likely to shop locally daily compared to people from an advanced age. This is because older people are retired, they usually have more free time than younger people who are studying or working. As the international centre on ageing says, municipal markets survive because of the elderly (Centro Internacional sobre el Envejecimiento, 2018.) If we analyse the motivators for purchases (sub-objective 2.1), we can affirm that the main reason for young people to buy in local shops in Villena is trust in the shopkeepers (48.5%). Time is also given great importance, with many of the surveyed (30.5%) citing travel savings as a reason. As mentioned above, young people have less free time to go shopping so saving time by purchasing products on shops located in the town is key for them. This statement is further affirmed by the fact that in another question, people between 18 and 44 years of age stated that they shop at Villena market because of its location and proximity. Age also influences where each demographic group shops. Due to a greater understanding of technology, youth make greater use of digital platforms than older people to make purchases. Justifying this with data, 67.5% of young people shop online while only 58.7% of older people do so. Young people show a higher frequency of shopping in ecommerce, with 36.7% doing so on a weekly or monthly basis. This result drops to half, 15.9%, among people over 44 years old. Another relevant place for young people to shop is in shopping centres, to the detriment of municipal market and Thursday Street market (61%). Younger shoppers are gradually moving away from traditional commercial establishments. A vast majority of the young people, 89,3%, that visit the municipal marketplace, purchase fruits and vegetables. In addition, nuts (69,8%) and meat (51,5%) are other relevant are other categories that have a considerable number of customers. Younger people are less likely than older people to buy canned food. On the other hand, they show a greater interest in prepared food, with 21.9% doing so compared to 10.3% of the older population. This Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 33 demonstrates that purchasing preferences for product categories among different age demographic groups exist. In order to get a visual picture of what people between 18and 44years old buy in the local market in Villena, the following graph has been prepared which presents the results in aggregated categories. Figure 22: Type of purchase do you usually make do young people make at Villena urban commerce Source: compiled by author based on obtained data Young people are satisfied with local commerce in Villena, with a score of 4.33 out of 5. Referring to the category of leisure, dissatisfaction among young people grows as 12.8% of those who buy this category rate it with 1 or 2 stars. The problem can be linked to the fact that 57,3% of the youth do not find what they are searching for in the commerce of the city. This is a point on which both the Villena council and ACV should work, so that the leisure demands of the inhabitants are covered. 5.3. MULTIVARIATE ANALYSIS In this phase of the work, a multivariate analysis will be undertaken using the cluster technique. This aims to understand the differences in the behaviours of groups that are different from each other and how people who are similar act. 5.3.1. Characterisation of consumers in Villena by profiling A characterisation of consumers in local commerce in Villena will be conducted by performing a cluster analysis. The Ward method will be used as it will allow to group the variables, achieving the greatest homogeneity and the greatest divergence between them (De la Fuente, n.d.). This will help into responding sub-objective 1.4. For the characterization of buyers, the sample is differentiated according to a list of variables that differentiate the respondents without identifying them (this distinction is important, as will be seen). The differentiating variables for classification by cluster analysis are presented in the table below. These variables have been chosen as they are the original ones and are the ones that best summarize who are the satisfied and unsatisfied shoppers. The variable name has been translated to respect the language in which the paper is being written. 90,80% 68,50% 49,80% 36,90% 35,30% Groceries Fashion/complements Beauty/health Leisure Home/decoration Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 34 Table 4: Differentiating variables for classification by cluster analysis Shop_Groceries What type of purchase do you usually make? (Groceries) {Yes/No} Shop_Beau_Health What type of purchase do you usually make? (Beauty/ Health) {Yes/No} Shop_Home_Deco What type of purchase do you usually make? (Home/ Decoration) {Yes/No} Shop_Leisure What type of purchase do you usually make? (Leisure) {Yes/No} Shop_Fashion What type of purchase do you usually make? (Fashion/ Complements) {Yes/No} SatisfLC_Dico Are you satisfied with local commerce? {Yes/No} Source: compiled by author based on obtained data By means of a cluster analysis, two segments have been defined and characterized by means of these variables. The individuals could have been grouped into four groups, achieving a clearer differentiation, but to meet our objective, a segmentation into two groups was more convenient. To do this, cross-tabulations of each of the above variables with respect to the created variable Clu_2: {Cl1; Cl2} have been used. A comparison of the distribution of the proportion of "Yes" versus "No" responses in each cluster are used. The following table summarizes this comparison in each cluster (For more detailed information, see the annexed tables on annex 3). Table 5: Summary of the comparison in each cluster Variable Interpretation Shop_Groceries Proportion of "Yes" over " No": Slightly higher for Cl2. Shop_Beau_Health Proportion of "Yes" over " No": Much higher for Cl2. Shop_Home_Deco Proportion of "Yes" over " No": Much higher for Cl2. Shop_Leisure Proportion of "Yes" over " No": Much higher for Cl2. Shop_Fashion Proportion of "Yes" over " No": Much higher for Cl2. SatisfLC_Dico Proportion of "Yes" versus "No or Not so satisfied with local commerce": Slightly higher for Cl2. Source: compiled by author based on obtained data As can be seen from the table above, in general Cl2 is more likely to contain individuals who shop in all commerce categories, and who are generally very satisfied with local commerce. Therefore, CL2 will be referred to hereafter as the Enthusiasts segment, and in contrast, Cl1 is referred to as the Reluctant. A closer inspection of the data in the annexes section (annex 3) shows that the categorizing variable is Home/Decoration as all the people who shop in this category are classified as Enthusiasts. The Beauty/Health, Leisure and Fashion/Complements variables reinforce this as they help to classify the clusters. The graphs and data indicate, in relative terms, how Enthusiasts contain a higher number of people who shop in these categories than people who do not. With this differentiation in mind, the next step is to further characterize and identify the components of each segment by considering an even larger list of variables. By crossing the two clusters with the socio-demographic variables, it will be possible to compare and describe how both Enthusiasts and Reluctant buyers are. The following variables are used for this purpose: Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 35 Table 6: Characterization of clusters with sociodemographic variables Variable Interpretation Age Age groups (4): Ratio of Reluctant (CL1) over Enthusiasts (CL2). 18-34: Much higher (Cl1 >> Cl2). 35-44: Higher (Cl1 > Cl2). 45-54: Higher (Cl1 > Cl2). 55 or more: Slightly higher (Cl1 < Cl2). Gender No differences. Familiar situation No differences. Residence No differences. Source: compiled by author based on obtained data According to the results of the comparison of response proportions, it can be observed that only the Age variable can provide some insight into how each cluster is characterized. The proportion of Reluctant respondents is much higher among the youngest, and higher in the next two age groups. Finally, among the oldest (55 or more), the two groups are almost equal, with some slightly higher proportions of Enthusiasts than Reluctant respondents. The results can be observed more extensively in Annex 4 where each cross table explain each of the variables above. The other socio-demographic variables (gender, family situation and residence) do not show significant differences to be able to categorise the clusters, it is therefore assumed that there is no association between these variables and the segmented groups. If the graphs in the annexes are observed, it can be seen that they show similar composition. Summarizing, it can be said that Age is the categorising variable of the two clusters. Within the Reluctant segment there is a greater presence of young people, while the older ones tend to be the Enthusiasts. Finally, the following relation of variables is examined with respect to the proportion of Reluctant and Enthusiasts. It is important to note that none of these variables were considered in the creation of the classification rule, so if there are significant differences, this is a sign that the differentiation between Enthusiasts and Reluctant goes beyond the variables cited in the first table. Table 7: Characterization of clusters with more variables Variable Interpretation How often do you shop at local shops in Villena? No differences. Do you usually shop at the municipal marketplace? Yes: higher presence of Enthusiasts. No: higher presence of Reluctant shoppers. Do you usually buy in Thursday’s Street market? Yes: Relatively larger presence of Enthusiasts. No: Relatively higher presence of Reluctant shoppers. Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 36 Do you shop in online commerce? No: Relatively higher presence of Reluctant shoppers. Yes: Majority choice in both groups, although with relatively higher presence of Enthusiasts. Do you shop in shopping malls? No differences. Source: compiled by author based on obtained data If the characterisation variables are analysed individually, it can be seen that the variable measuring the frequency of purchases in Villena's shops does not give any differences, since in relative terms, the frequencies are similar. Therefore, there is no association between frequency and belonging to one or other cluster. This is also repeated with purchasing in shopping malls, so there is also no association between going to shopping malls and being Enthusiastic or Reluctant. Referring to the cross-tabulation of whether individuals from the different clusters shop at the municipal market, the following can be stated. There is an association between shopping at the municipal market and belonging to one cluster or another. This is because in relative terms, there is a greater presence of individuals categorised as Reluctant (47%) than Enthusiasts (33.6%) who do not shop there. This association is also observed with shopping or not shopping at Thursday's Street market. Among those who do not shop there Reluctant are, in relative terms, predominant with 44.6%. Also, in relative terms, there is a higher number of Reluctant (43.2%) who do not buy from e-commerce than Enthusiasts who do not buy from ecommerce (26.2%). Tables can be reached on annex 5. 5.4. ANALYSIS OF OPEN-ENDED QUESTIONS As mentioned in the description part of the survey, part 7 of the questionnaire was dedicated to finding out people's perceptions and opinions about commerce in Villena. In this way, respondents were asked which products or brands they missed in Villena, giving them the opportunity to write a free response. This analysis will help giving an answer to subobjective 3.2. A total of 559 responses were obtained, all of which differed from each other. To be able to analyse the responses, they needed to be categorized. To do this, several categories and sub-categories had to be created and classified response by response. As can be seen in the following table, the main answer categories were fashion, none, leisure, promotions, and others. Table 8: Categorization of the open-ended questions Categories and sub-categories Number of answers Fashion 217 General 56 Specific shops 48 Children 41 Sports 22 Youth 22 Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 37 Others 16 Large sizes 12 None 130 None 130 Leisure 99 Youth 26 General 21 Shopping mall 20 Electronics 16 Cinema 10 Others 6 Promotion 71 Variety 47 Price 9 Service 6 Marketing 6 Quality 3 Others 42 Hostelry 16 Trends 13 Home 7 Others 6 Total of answers 559 Source: compiled by author based on obtained data As it can be seen in table 3, a variety of responses were obtained from the open-ended question. The most repeated category is “fashion”, as 38,8% of the surveyed are missing products from this type of category in Villena's shops. Leaving aside people who miss fashion products in general, it is worth mentioning that 16% of people have a special interest in children's and youth clothing. There is also a huge interest in specific brands such as Primark or shops like Zara, Bershka and Pull & Bear from the Inditex group. An interesting note within this category was the interest shown by 12 people in large sized clothing. Whether the lack of specific shops or products for special groups of people, this category is the one that creates the most conflict in the village. Secondly, 23,26% of those surveyed indicated that they had not missed any product in the local shops in Villena. In fact, they were very happy as they find every product they are looking for. Some of them also indicated being satisfied with the town's offer, as it was in line with their capacity. “Leisure” was the third most mentioned category in the responses (17,71%) since many people were dissatisfied with the town's offer in this area. There is a demand for more leisure-related activities for the younger demographic groups. On the other hand, it is worth mentioning that a considerable part of the respondents demand infrastructure such as shopping malls and cinemas. In fourth place, with 12,7% of the responses, is the category of “promotions”. In this category, responses concerning the lack of variety have been added, which account for 8.4% of the total. In smaller percentages, people are missing lower prices, better product quality and a higher quality of service offered by the shops. Mention should be made here of the lack Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 38 of promotion of activities and shops and the fact that in Villena, when returning products, the money is not refunded but vouchers are given back. In the “others” category, the responses that could not be added in the other categories were collected. 16 people showed a special interest in improving the range of restaurants in Villena, with special mention of fast-food chains such as MacDonalds. Finally, the lack of trendy products such as vegan and healthy products, which are on the lips of most people, was also mentioned. 6. CONCLUSIONS, DISCUSSION, RECOMMENDATIONS AND LIMITATIONS The objective of the last stage of this investigation is to present the conclusions that are obtained with the analysis, as well as opening a debate about the topic. Additionally, recommendations from the obtained results will be provided to Asociación de Comerciantes de Villena and the town council so that they can improve their urban local shopping experience. 6.1. CONCLUSION Customer shopping habits have transformed over the last decades due to changes in the tastes and preferences of the population. We live in a digital era in which many of the purchases are made through digital channels. In addition, along with this, healthy and environmentally friendly lifestyles have been established, so managers must try to offer products and services that meet these preferences. The COVID-19 pandemic has directly affected the economy and urban commerce which has influenced on managers having to restructure businesses and strategies. In order to understand and study human behaviour and the conditioning factors of the shopping trends on a medium sized town/city in Spain, a quantitative investigation of the market was carried out between Neuromobile, Jon Charterina and the author of the study, Julen Pérez. This has made possible to draw a series of conclusions in relation to the proposed objectives. The survey has enabled to make a description of the behaviour of individuals regarding local commerce in Villena (objective 1). In relation to this and sub-objective 1.1, it was found that people mainly buy in local shops on a weekly and daily basis. This is due to the fact that the main category of product bought in Villena is groceries, which is a product with a high turnover (sub-objective 1.2). Deeper data regarding shopping frequency and product preferences can be found at point 5.1. Local shops are losing their younger customers due to their changing tastes and shopping preferences, and therefore the parties involved in the project proposed the objective of understanding how this group behaves (sub-objective 1.3). They are increasingly moving away from traditional commerce, as 67.5% of young people buy through online channels and do so much more frequently than older people. They also show a greater interest in novelty products, such as prepared food. This group, as is to be expected, is the one that shows the greatest dissatisfaction with leisure in Villena. Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 39 The cluster analysis led to the conclusion that there are two groups of shoppers in Villena's local shops, the Enthusiasts and the Reluctant (sub-objective 1.4). The Reluctant are characterised by the fact that they are the youngest shoppers and that they do not buy products in the home/decoration category. They also tend not to shop at the Thursday's Street market or at the municipal marketplace. Enthusiasts, on the other hand, are the older shoppers, aged 55 and over, who shop in almost all product categories and go to both the municipal marketplace and Thursday's Street market. Objective 2 aimed to find the factors that act as leverage and motivate people to shop locally. In this sense, people indicated that the trust they have with the salespeople has a great influence because they allow them to see the products up close, as well as to try them out and touch them. The second most named factor was the time saving, as the proximity of the shops avoids people having to drive to other places to do their shopping. Regarding sub-objective 2.1, the reasons why individuals shop in Villena's shops do not vary according to age, as the frequency of responses is similar between young and old people. There is only a slight difference in that older people buy in local shops because of the treatment and trust in the seller. The survey concluded that people in Villena are satisfied with local commerce (objective 3). This is because the commerce was rated 4.36 out of 5 points. However, if we look at the problems that people see in Villena (sub-objective 3.1) we can see that they indicate that sometimes they do not find the products they are looking for. They also say that it is not easy to be informed about promotions and that the products are not only expensive, but also that the shops often do not stock the most up-to-date brands. Regarding subobjective 3.2, it has been possible to obtain opinions about the products that the people who shop in Villena most miss. Based on the fact that only 23,26% of people are totally satisfied with the offer, the rest demand a greater variety of products and services. The people surveyed particularly emphasize the lack of fashion and leisure products and services, especially for the younger age groups. 6.2. DISCUSSION AND RECOMMENDATIONS After completing the qualitative research and analysing the results obtained, it is possible to deduce a series of important factors in the field of urban commerce that the association together with small businesses should take into consideration to improve shopping experience and the services they provide to citizens. This would answer the fourth and last objective of the research. The most obvious conclusion from this paper is that young people are distancing themselves from local commerce in the town as they are more reluctant to shop. Therefore, both the association and the shopkeepers must look for ways to encourage them to stop being reluctant and become enthusiastic, or at least to do more shopping in Villena. In this sense, it was also found that young people are the most dissatisfied with the fashion products and accessories and with the leisure activities, which are insufficient in the town. In particular, in the open-ended question, mention was made of the lack of shops such as Zara or Primark. Opening shops belonging to these companies would not be convenient for small businesses, as small local clothing shops cannot afford to set the low prices of the first Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 40 ones. Furthermore, as the purpose of the study is to improve the local shopping experience in Villena, making such a recommendation would go against the objective. Therefore, my recommendation is based on trying to offer fashion products in a non-traditional way, i.e. to encourage the purchase of fashion from local shops through the experience. One of the open responses analysed, gave me the answer on how implement the recommendation, through second-hand markets. My experiences and stays in other countries have allowed me to discover that this activity is very well established abroad. In particular, I was surprised to see how many secondhand clothes markets and shops there are in Helsinki, and how people get together and spend their afternoons looking for clothes. But this trend towards vintage clothing is also a reality in cities like Madrid and Barcelona. Among the benefits, it is worth highlighting that the clothes that are bought tend to have lower prices. In addition, it contributes to a lower consumption and production of clothes, as the clothes that are no longer worn are reused. On the other hand, I would recommend holding an annual alternative market, in which the clothing, footwear and accessory shops of Villena would offer their products in an outdoor market. In this way, fashion, design and recycling would be promoted, breaking away from the everyday. Both events or activities would help both to satisfy the fashion needs of the town and to promote leisure, as they are a good excuse to get together with family or friends and spend a good morning or afternoon. Another way to improve the shopping experience in Villena is to look at the reasons why individuals shop there. As can be concluded, time saving is a relevant factor when it comes to shopping in Villena's local shops. Taking this point into account, it would be possible to go further and set up a service to deliver products from local shops to customers' homes. Dendak Bai is the association of shops in Durango, the town where I live, and has already set up this type of service. It works in such a way that the customer fills in a form indicating the details and the establishment has to contact the delivery drivers, so that they can make the delivery to the place and time desired by the customer. The shop charges the customer the price set for the delivery when the customer makes the payment to the shop. At the end of the month, the shop pays the association for all delivery services. The addition of this service in Villena's local commerce would make shops and retailers more flexible, thus offering greater value to customers. Lastly, it is worth mentioning that data-driven management is a reality today and must be executed correctly. As Marcos García 4 stated in one of the many meetings that have been held to carry out this study, gone are the days when managers of small traditional stores made their decisions through feelings and perceptions. A step forward must be taken and go with the flow of digital transformation and managing from data is a pending task. More important than the quantity of data is the quality of the data, because if the data is not descriptive and relevant, it will not allow designing strategies for change. Once the conclusions are drawn, strategies should be set to solve the problems that the commerce is facing and I recommend benchmarking actions. In this way, strategies that have been successful in other cities or towns could be implemented. It is at this point where specialized companies such as Neuromobile must align with associations and businesses in order to get the most out of it and row in the desired direction. 4 García, M. Op. Cit. <24/02/2022> Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 41 6.3. LIMITATIONS AND FUTURE LINES OF INVESTIGATION The market study that has been carried out has a series of limitations that affect somehow the efficiency and accuracy of the data and therefore the results obtained. In the first place, the survey was designed and carried out to obtain data about commerce in Villena. Even though it is a medium-sized city, it is not possible to make decisions aimed at other cities or towns of similar size based on the results obtained from the survey. It can give us a small brushstroke to make recommendations to other cities, but the ideal would be a study of the area since in Spain, not all regions and provinces have the same habits, purchasing power, commerce or quality of services. Secondly, there is many participants, 157 to be exact, who did not want to disclose their gender, ticking the "Other" box. This is an important figure as it is 28.1% of the total number of respondents, and even more knowing that only 73 of the participants are men. If the sex of these people had been known, the statistics obtained by segmenting by sex would have been much more accurate, and the presence of men would surely increase. Using predictive methods, the company in charge of the survey offered to deduce the sex of these people, but in the end, it was chosen not to do so. Another constraint was the fact that the survey had already been conducted. Initially, a great effort had to be made to understand why and for what purpose the questions in the survey had been chosen. The design of the survey meant that many responses were lost along the way, because if, for example, respondents indicated that they did not buy leisure products, satisfaction with that category could not be assessed. At first glance, this may not seem like a limitation, but when we tried to perform logistic regression functions, they did not give the expected results. It is true that 559 is a considerable number of individuals, but a larger participation would have helped to obtain more reliable analyses. With a view to future research, it would be interesting to conduct a second survey after some time and once some of the recommendations made in this work have been implemented. If young people's perceptions of local commerce in Villena had improved, it would be a sign that the recommendations made by the author are accurate and correct. Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 48 Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 49 Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 50 Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 51 Annex 3: Differentiating variables for classification by cluster analysis Cross table: What type of purchase do you usually make? (Groceries) * Ward Method Ward Method Total 1 2 What type of purchase do you usually make? (Groceries) No Count 36 2 38 Expected count 23,5 14,5 38,0 Yes Count 309 212 521 Expected count 321,5 199,5 521,0 Total Count 345 214 559 Expected count 345,0 214,0 559,0 Cross table: What type of purchase do you usually make? (Beauty/ Health) * Ward Method Ward Method Total 1 2 What type of purchase do you usually make? (Beauty/ Health) No Count 208 47 255 Expected Count 157,4 97,6 255,0 Yes Count 137 167 304 Expected count 187,6 116,4 304,0 Total Count 345 214 559 Expected count 345,0 214,0 559,0 Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 52 Cross table: What type of purchase do you usually make? (Home/ Decoration) * Ward Method Ward Method Total 1 2 What type of purchase do you usually make? (Home/ Decoration) No Count 340 0 340 Expected count 209,8 130,2 340,0 Yes Count 5 214 219 Expected count 135,2 83,8 219,0 Total Count 345 214 559 Expected count 345,0 214,0 559,0 Cross table: What type of purchase do you usually make? (Leisure) * Ward Method Ward Method Total 1 2 What type of purchase do you usually make? (Leisure) No Count 258 83 341 Expected count 210,5 130,5 341,0 Yes Count 87 131 218 Expected count 134,5 83,5 218,0 Total Count 345 214 559 Expected count 345,0 214,0 559,0 Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 53 Cross table: What type of purchase do you usually make? (Fashion/ Complements) * Ward Method Ward Method Total 1 2 What type of purchase do you usually make? (Fashion/ Complements) No Count 141 40 181 Expected count 111,7 69,3 181,0 Yes Count 204 174 378 Expected count 233,3 144,7 378,0 Total Count 345 214 559 Expected count 345,0 214,0 559,0 Cross table: Are you satisfied with local commerce? * Ward Method Ward Method Total 1 2 Are you satisfied with local commerce? Yes Count 174 126 300 Expected count 185,2 114,8 300,0 No or not so much Count 171 88 259 Expected count 159,8 99,2 259,0 Total Count 345 214 559 Expected count 345,0 214,0 559,0 Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 54 Annex 4: Cross tables between clusters and socio-demographic variables Cross table: Age groups * Ward Method Ward Method Total 1 2 Age groups (4) 18-34 Count 73 34 107 Expected count 66,0 41,0 107,0 35-44 Count 121 67 188 Expected count 116,0 72,0 188,0 45-54 Count 98 55 153 Expected count 94,4 58,6 153,0 55 or more Count 53 58 111 Expected count 68,5 42,5 111,0 Total Count 345 214 559 Expected count 345,0 214,0 559,0 Cross table: Where do you live? * Ward Method Ward Method Total 1 2 Where do you live? Villena Count 323 201 524 Expected count 323,4 200,6 524,0 Another town or city Count 22 13 35 Expected count 21,6 13,4 35,0 Total Count 345 214 559 Expected count 345,0 214,0 559,0 Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 55 Cross table: Gender * Ward Method Ward Method Total 1 2 Gender Male Count 42 31 73 Expected count 45,1 27,9 73,0 Female Count 211 118 329 Expected count 203,1 125,9 329,0 Other Count 92 65 157 Expected count 96,9 60,1 157,0 Total Count 345 214 559 Expected count 345,0 214,0 559,0 Cross table: Familiar situation * Ward Method Ward Method Total 1 2 Familiar situation Single Count 96 50 146 Expected count 90,0 56,0 146,0 Couple Count 60 42 102 Expected count 62,9 39,1 102,0 Family Count 180 117 297 Expected count 183,1 113,9 297,0 Total Count 336 209 545 Expected count 336,0 209,0 545,0 Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 56 Annex 5: Cross tables of variables used for cluster characterization Cross table: How often do you shop at local shops in Villena? Total Once a month or more Less than once a month Ward Method 1 Count 309 36 345 % Within Ward Method 89,6% 10,4% 100,0% % Within How often do you shop at local shops in Villena? 60,8% 70,6% 61,7% 2 Count 199 15 214 % Within Ward Method 93,0% 7,0% 100,0% % Within How often do you shop at local shops in Villena? 39,2% 29,4% 38,3% Total Count 508 51 559 % Within Ward Method 90,9% 9,1% 100,0% % Within How often do you shop at local shops in Villena? 100,0% 100,0% 100,0% Cross table: Do you usually shop at the municipal marketplace? Total No Yes Ward Method 1 Count 162 183 345 % Within Ward Method 47,0% 53,0% 100,0% % Within Do you usually shop at the municipal marketplace? 69,2% 56,3% 61,7% 2 Count 72 142 214 % Within Ward Method 33,6% 66,4% 100,0% % Within Do you usually shop at the municipal marketplace? 30,8% 43,7% 38,3% Total Count 234 325 559 % Within Ward Method 41,9% 58,1% 100,0% % Within Do you usually shop at the municipal marketplace? 100,0% 100,0% 100,0% Julen Pérez Blanco: Data-driven urban commerce management: The case of Villena 57 Cross table: Do you usually buy in Thursday’s Street market? Total Yes No Ward Method 1 Count 191 154 345 % Within Ward Method 55,4% 44,6% 100,0% % Within Do you usually buy in Thursday’s Street market? 56,3% 70,0% 61,7% 2 Count 148 66 214 % Within Ward Method 69,2% 30,8% 100,0% % Within Do you usually buy in Thursday’s Street market? 43,7% 30,0% 38,3% Total Count 339 220 559 % Within Ward Method 60,6% 39,4% 100,0% % Within Do you usually buy in Thursday’s Street market? 100,0% 100,0% 100,0% Cross table: Do you shop in online commerce? Total No Yes Ward Method 1 Count 149 196 345 % Within Ward Method 43,2% 56,8% 100,0% % Within Do you shop in online commerce? 72,7% 55,4% 61,7% 2 Count 56 158 214