Brazil World Cup: “Travelmetrics”, or how data crunching could defy conventional wisdom

Pascal Clement

Head of Travel Intelligence, Amadeus IT Group

This content is only available in this language.

If you watched Brad Pitt’s film Moneyball (2011), you know the basics about sabermetrics. If you didn’t, the brief description out of Wikipedia might help you: “empirical analysis of baseball, especially baseball statistics that measure in-game activity”. This could actually be translated into a bunch of smart data scientists crunching tons of data and trying to generate “advanced statistics” in order to get a better and deeper knowledge of the game of baseball. In the film, Brad Pitt uses sabermetrics to defy conventional wisdom and guide a group of cast offs and undervalued, cheap players to a record streak of victories and even championship contention. In short, he uses deep understanding of his business through statistics to exploit market inefficiencies.

So where does Amadeus fit in this story, you might be asking yourself? Well, although there is no sign of Brad Pitt wandering in our premises, we do have the data and the data crunching skills. And we don’t use them for the all-important goal of winning the World Series of baseball, but rather to transform raw data into meaningful information, by generating “advanced statistics” in order to get a better and deeper knowledge of the travel industry. This could help defy conventional wisdom and empower our partners to exploit market inefficiencies. Want some proof? Let’s have a look at the World Cup, to be hosted in Brazil during June and July this year.

The Brazil World Cup in figures

I am pretty sure that you have heard the traditional claim by big events organisers that tourism flow into the country will grow substantially. Embratur – the Brazilian Tourism Institute – last year estimated that around 600,000 foreign touristswill visit Brazil for the World Cup. Conventional wisdom says it will probably be the case. Even statistics from past World Cups back these estimations: South Africa achieved a 22% increase in tourism earnings after the 2010 Football World Cup.

Let’s see if our “travelmetrics” can give us a deeper, more precise look at those predictions? We can start by crunching some search data. As you can see in the graph below, searches for travel to Brazil during June and July peaked on December 6 2013. This, not coincidentally, is the day that the FIFA held the tournament’s final draw. Since then, search volumes have remained relatively high.


So we now know there were more searches for flights to Brazil, but we need to make sure these were really driven by the World Cup. The graph below confirms the previous premise as when we looked at the departure days entered in the searches performed, we see is that searches for departures in the days around June 12 are more than 10 times higher than in May or August. Not a surprise, again, given that the tournament’s kick-off will take place on that date. From this perspective, it seems that Embratur’s predictions might be correct.

But again “travelmetrics” are all about deeper looks at numbers. So let’s take our analysis a step further. One of the best proxies to know if that search – which can be seen as potential demand – will really transform into a bigger tourism flow is, of course, capacity increase in scheduled flights to Brazil. Capacity for June and July will go up by 7% and 5% respectively. This seems like a notable increase, especially when compared to last year’s 1% and 3% hikes… but not so much when you realise that it is just about the average growth for the year, and that August will see an 8% increase in seats available. So maybe those potential 600,000 new foreign tourists will have it difficult getting to Brazil?

Missed Opportunity?

worldcup most searched-for City pairs

It might as well be because of market inefficiencies – remember, “travelmetrics”, as sabermetrics, are best used at finding market inefficiencies. When comparing search and seat capacity data, the most striking finding is that 47% of searches have Rio de Janeiro as a destination, but seats available for Rio are less than half when compared to those available for Sao Paulo, which only gets around half the searches.

So does this mean airlinesare missing on a business opportunity by not offering enough flights to Rio? Or that, against conventional wisdom, the World Cup will not bring the promised massive flow of new travelers? Of course this is just a light approach to the full scope of Travel Intelligence, and in order to get #8226etproof conclusions we would need to include a multitude of further factors, such as real bookings and seat-load factor, a few of which are not yet available.

But it serves as a sneak-peak to the ability of “travelmetrics” to dig further into data. And of course with a little creativity, we can even use them to know which fans are more confident on their team’s capability to bring the World Cup back home with them…but you will have to wait a few weeks to find out more.



Latin America, Big Data, Research