We are all quite aware of the meaning of smart city, that is, of using digital technologies and, more generally, technological innovation, to optimize and improve infrastructure and services to citizens, making them more efficient.
Not only thinking of resident citizens, but also temporary citizens, such as travelers, tourists, and today also digital nomads and smart workers.
However, the concept of smart city cannot be fully implemented, in my opinion, if not through the intelligent use of data to objectively understand what happened in the past, what is happening right now, more or less in real-time, and, above all, what could happen in the future.
The concept of sensors, cameras, and IoT (Internet of Things) is implicitly linked to the idea of smart city, to have environmental data and surveys of the city or territory to be monitored. However, today, more easily implemented solutions have less friction and technological complexity and are more economically sustainable.
It becomes more accessible and easier to use available data sources even without installing thousands of sensors, such as online data and satellite data. It is, therefore, necessary to find interpretative models that extract value from these data by linking “direct” information (the CO2 value or traffic, for example, which are both available from satellite sources) with “proxy” or correlated values, such as the hotels or airline tickets rates, modulated according to the dynamics of the destination and the demand.
Artificial intelligence simplifies this process allowing information to be extracted from so-called big and alternative data (social media data, for example) and relating them to more traditional data sources.
Artificial intelligence is also valuable when defining predictive models which, by analyzing the historical series, and weak signals within hundreds of different sources, allow a projection – or a forecast – of some future phenomena.
This is what we do at The Data Appeal Company: the collection and analysis of all online data sources, integrated with contextual data on the destination, such as events, weather, hotel rates, and OTA saturation.
These online data are closely related to physical phenomena in the area, as the volume of digital content is directly proportional to the number of people present on the site itself. The correlation between the volume of online data and, for example, the official ISTAT flows (i.e., Italian statistic organization) is practically perfect. This allows to have reasonable confidence in the reliability of the results even in real-time, and above all, it will enable to build predictive models about the future.
These are the data we analyze to understand a destination:
- data from Online Maps;
- perception data (reviews, comments, social posts, photos, etc.);
- contextual data (weather, events, prices, expense data, mobility in the area, etc.);
- travel intentions data: future research and bookings. (flights and accommodation).
The real-time reading of these data and their correlations allows us to represent what is happening in the area and predict precisely how some phenomena will evolve in the future.
Data-driven destination: the use of data
The data highlights the phenomena of the destination with greater objectivity and makes it possible to give the right priorities and make more informed decisions. Which areas are perceived to be safest? The most popular ones? Which are the most relevant attractors of the destination according to each market? How do the French, the Americans, or the Germans move? Which hotels do they choose, which restaurants or museums, and in which areas of the destination?
Here is how the data help to understand visitors’ choices and their reasons with greater objectivity.
It is the consequence of the strategic definition illustrated above. In this case, we try to understand if the strategy has been well designed and/or effectively implemented, allowing us to correct it on the go if something does not work.
Defining and planning marketing operations in an objective and data-based manner allows one to focus efforts, optimize budget, minimize risks, and monitor the results of campaigns during actual planning to correct any deviations.
Furthermore, by semantically analyzing the contents, it is possible to understand travelers’ perception of certain markets and modulate promotional messages to specific sensitivities by building personas in a data-driven way and not only based on personal assumptions and intuitions.
Having predictive models allowing to evaluate future scenarios enables to plan resources to improve the tourists’ experience and, at the same time, reduce the impact of flows on resident citizens.
Private operators can plan work shifts, resources, and procurement based on precise flows, even pushing themselves in the choice of purchases and staff tailored to specific markets, their language, and tastes. That is, trivializing, if the data show a significant presence of Germans in a given weekend, maybe you can make sure that you have German speakers on duty and prepare the service by ordering those products, which from the analysis of the data are usually appreciated by the Germans, and so on.
The governance of the destination could enhance public transport, alert the traffic police and perhaps make sure that the street cleaning, garbage collection, and emptying of bins service is fully operational even during that weekend, as well as all the tourist information services.
These are trivializations and elementary examples that immediately give the idea of how intelligent use of technology, data, and predictive models can be one of the ways to reconcile the dimension of resident and temporary citizens before falling again into the controversy and discontent over what was called the pre-pandemic overtourism.