Big Data to Enhance the Tourism in Mexico

Juan Murillo Historias con datos

Last November 29th we presented the results of the project “Big Data and Tourism” carried out together with BBVA Bancomer and the Secretariat of Tourism in Mexico (SECTUR).

Due to its cultural heritage and to the natural diversity of the country, among other strengths, tourism in Mexico is a rising sector, a solid source of employment in the country, and a relevant growth pillar that contributed in 8,7% to the GDP in 2015.

Leveraging on the descriptive capacity of financial data, the aim of our partnership project was to create new insights based on the spending patterns of national and international visitors. The driving features of the collaborations have been the following:

About the data: We employed card payments data, a high resolution data source throughout time and space that, besides, reflects real facts, not just intentions or recalls. Its statistical value has been demonstrated in other projects through the replication of macroeconomic indexes. It has been a firm intent for innovation from SECTUR and INEGI to rely on big data sources in the frame of this pioneering project.

About the methodology: Whereas foreign tourism flows are relatively easy to track and measure, given that the entry points to the country are limited, measuring national inner tourism flows is a much harder task. Dealing with anonymized data, we have developed and implemented a new model to assess the usual environment for everyone of Bancomer cardholders, so every time a transaction is made out of a card tenant usual environment, it is labeled as a transaction linked to touristic activity, and is aggregated with other transactions to be part of the statistical results. This has allowed us to characterize three different kind of profiles: residents, national visitors, and foreign tourists.

About the infrastructure and tools: Once anonymized, data are stored in a cloud infrastructure, where processing and analysis are carried out. This is a much more economical and flexible solution than on premise infrastructures. Regarding the tools employed, for the data analysis mainly R software has been applied, and to visualize results, Tableau and CARTO have been used.

About the team: Data Science is a new discipline that combines advanced skills from STEM education (science, technology, engineering and mathematics). BBVA has concentrated this talent on its analytic centre of excellence -BBVA Data&Analytics- but, thanks to internal training programmes and to the activity of its Community of Practice on Data Science, this knowledge is spreading throughout the organization as a vector of the digital transformation. In the development of this project, an interdisciplinary team in Mexico and Spain has been involved, being Juan de Dios Romero, Heribert Valero, Beatriz Alonso, and Juan Murillo the members on BBVA Data&Analytics side.

About the results: the memorandum of understanding signed between SECTUR and Bancomer in October 2015, comprised the development of 12 analysis throughout 2016. The presentation made to media and stakeholders last November 30th focused on first two deliverables: “Magical Towns” and “Quintana Roo”.

Our results reveal important aspects such:

  • Where does a greater concentration of tourism arise? Thanks to the usual environment assignation to each of every 20 million Bancomer customers, we have been able to measure the relative importance of national visitors, international tourists, and local inhabitants, showing spatial distribution of these three segments, and its evolution throughout time, being able to measure seasonality for every destination in very accurate way.
  • Which are the activities or commercial categories that present the highest expenditure?  we have provided average ticket, a metric that varies among different destinations and profiles, for each collective (national and international tourists), and also their time variations and trends.
  • Which is the country of origin of the foreign tourist, or what state does the national tourist come from? This is useful to make a ranking between the global tourism issuing markets.

As main conclusions in the Quintana Roo area, we can highlight that 35% of the expenditure made by foreign tourists was carried in Playa del Carmen, while the national tourist spending takes special relevancy in Cancún and Playa del Carmen, with 25% and 27% of expenditure respectively. Likewise, the highest concentration of tourists was registered in the periods of July/August, and November/December, and the categories with more relevance were “entertainment” for the foreign visitors and “travels” for national tourist. Around 50% of the expenses incurred by national tourists were made by Federal District, State of Mexico and Quintana Roo inhabitants, in that order. In terms of foreign tourism, the major percentage of spending was generated by Americans tourists, with Argentina ranking second.

In general, among the destinations belonging to the “Magical Towns” program, there is a greater activity of the national tourists compared with the international ones. In this type of destinations, the activity of the national tourist is regional and with heterogeneous spending behaviors that depend on both geographical location and financial inclusion.

This type of knowledge allows the entities responsible for tourism in each country, like SECTUR in Mexico, to improve their offer, to target their promotional activity,  and thus increase the number of visitors and their experience.

A more detailed description of our approach is available (in Spanish) in this publication of “Magical Towns” Analysis and in this other publication of “Quintana Roo Analysis”.

We also invite you to look and share the visualization of the study.