New predictive models allow to show the evolution of tourists demand a year ahead.
Flight planning; agreements between hotels and tour operators to reserve places; hiring employees for the high season; food purchase… Every year, the tourism industry must make multiple decisions based on a simple premise: predict and guess what tourists’ behaviour will be. Artificial Intelligence and Big Data are taking action on the matter.
So far, the most used method to predict how the next season will go is to analyse flight searches or flight bookings for the next months.
“Being able to predict tourist dynamics is the great challenge. The problem is that this dynamic is very volatile, because unexpected things happen such as attacks, political convulsions, etc.”, explains Santi Camps, CEO and founder of the Spanish company Mabrian Technologies.
He adds however that “technology can help to understand these dynamics, predict changes and their consequences”.
Along these lines, Mabrian has developed a predictive model based on social behaviour indicators, that can show the evolution of tourist demand one year in advance.
The system is based on the constant monitoring of thousands of mentions on social networks (Instagram and Twitter) related to travel, and that takes into account the users’ profiles, even using facial recognition to deduce whether they are young or old.
“We can anticipate if a market will remain stable, increase or decrease for each destination”
This huge data volume is transformed into a series of indicators: security, hotel satisfaction, travel intention, etc. “With all these components, depending on how they vary, we can anticipate if a market will remain stable, increase or decrease for each destination”, says Santi Camps.
“Variations on security are key: if it remains stable, that part of the formula isn’t affected, but if there is a big drop, you can anticipate that bookings will be affected”.
This predictive model’s effectiveness has been proven in several cases. For example, in October 2017 it was anticipated that Egypt, Turkey and Tunisia would once again be tough competitors for Spain in the summer of 2018.
Indeed, Mabrian’s “Perceived Security Index” for these countries stabilised above 80 points (on a scale of 0 to 100), specifically among German tourists.
Another prime example is Cuba: thanks to Barack Obama and the restoration of diplomatic relations with the USA, in 2016 there was a significant increase in hotel prices, which was attributed to the great interest that Cuba was arousing among American tourists.
Some European tour operators considered whether it would be worthwhile to maintain their operations on the island if the very strong price escalation continued.
However, the USA market was generating 38% of the mentions on social networks about tourism in Cuba, but only 8% of hotel reviews. Canada, on the other hand, generated 48% of the hotel reviews.
At the same time, the intention to travel to Cuba from Italy, United Kingdom, Spain or Argentina had also increased.
“The conclusion was that travel demand from the US alone was unable to explain the rise in hotel prices and everything seemed to indicate that a call effect was being produced”.
In other words, tourists from other countries were rushing to travel to Cuba “before it changed” due to the supposed massive arrival of Americans and, therefore, that year’s price boom was only a temporary effect.
This article has been published in the September edition of the HOSTELTUR magazine.