What 3.5 Million Hostel Reviews Tell Us About the Next Top Travel Destinations

10 minute read

At TripHappy, we’re obsessed with using travel analytics to find where people are traveling to and where they are going next. We’ve previously taken a look at thousands of itineraries to find where the world travels and analyzed Reddit posts to find where Reddit travels.

But we started to wonder: how do the popularities of different travel destinations change over time as they become more well known and documented? Can we measure when a region, country, or city has become a “popular” destination? And can we use these findings to predict where the popular backpacker destinations will be next year and beyond?

To answer this question, we decided to take a look at all the reviews that have been submitted for properties on HostelWorld from 2012-2016, the world’s biggest hostel reservation platform. With this data, we should be able to measure trends in the growth of travelers at the city level.


We first started by aggregating all reviews at the property level and segmenting them by city, country, and region. This gave us a dataset of over 3.5 million reviews across 3,500 cities and 164 countries.

It was immediately easy to identify the most popular cities overall, with mostly Western European cities like London, Barcelona, Amsterdam, and Paris dominating the top 10. But we weren’t interested in finding the most popular places, but rather the destinations that have experienced explosive growth in a relatively short amount of time.

The reviews were then segmented by year and aggregated to the city level, giving us a matrix showing the numbers of reviews in a city by year. Given these actual review counts, we then determined the expected review counts based off of the total reviews in the city compared to the global number of reviews. This is similar to the procedure of running a Chi-Squared test for categorical data. The main difference here is we are not assuming independence between the year and the city - in fact, we're hoping to find depedence!

With these two matrices (one with actual review numbers, and one with expected), we then produce a third matrix by dividing the actual by the expected (A/E). Any city that's experiencing high growth will show A/E ratios greater than 1, especially in 2016 and 2015, indicating that more travelers have visited than would be expected if growth was constant. Similarly, any city with constant growth will have an A/E close to 1, and any city with negative growth will have an A/E below 0.

There's subtle wording here that should be clarified. An A/E ratio of 1, doesn't imply that a city isn't growing. Rather, that the growth rate is consistent with historic levels. For example, Montevideo, Uruguay experienced growth rates of 30% last year, but this is consistent with previous years and so it has an A/E ratio close to 1.


After running the analysis, here is our list of the top 10 most up-and-coming youth travel destinations

# City A/E Ratio % Growth (2016 vs. 2015)
1 Seminyak, Indonesia 2.34 223%
2 Gili Trawangan, Indonesia 2.27 253%
3 Da Lat, Vietnam 2.04 169%
4 Da Nang, Vietnam 1.96 194%
5 Yangon, Myanmar 1.89 148%
6 Ubud, Indonesia 1.87 195%
7 Pai, Thailand 1.87 165%
8 Mui Ne, Vietnam 1.87 171%
9 Salento, Colombia 1.86 203%
10 Luang Prabang, Laos 1.79 164%

The main story here is the humongous growth in Southeast Asia - the region went from a small blip on the radar to rivaling Western Europe in youth travel popularity. Some major cities like Bangkok, Chiang Mai, and Ho Chi Minh City were already gathering steam as early as 2008, so we don’t consider them to be “up-and-coming”. But the popularity of these cities has driven growth into other areas of the region, like Northern Thailand, the Thai Islands, Indonesia, Cambodia, Laos, and Central Vietnam. This “contagion” of popularity from a regional hub is something that we will explore in a future post.

Interestingly, as South East Asia has exploded in popularity, other countries in East Asia have stagnated or declined in terms of number of visitors. Countries like China (and Hong Kong), Taiwan, Korea and Japan have all lost steam as backpackers head South.

If we remove Southeast Asia from the candidate pool, our top 10 takes on a different characteristic

# City A/E Ration % Growth (vs. 2015)
1 Salento, Colombia 1.86 203%
2 Tulum, Mexico 1.61 190%
3 Medellin, Colombia 1.52 170%
4 Goa, India 1.49 175%
5 Cali, Colombia 1.43 148%
6 Cancun, Mexico 1.41 161%
7 Antigua, Guatemala 1.40 181%
8 Santa Marta, Colombia 1.38 151%
9 Jaipur, India 1.37 162%
10 Cartagena, Colombia 1.33 159%

Now we see Colombia stealing the spotlight with 5 cities, along with India, Mexico, and Guatemala. The massive growth in Colombia could be a result of the country's efforts to rid itself of the "Narcos" influence from the past decades. Especially with the recent FARC peace deal, Colombia could continue to see exponential growth in tourism.

We were surprised to see Goa and Jaipur on this list, considering their popularity has been known to travelers for many years. We believe this is a direct result of more Indian hostels coming online, and not necessarily a result of tourism growth. Similarly, for Mexico, we were surprised at Cancun's inclusion on the list. But, just like Colombia, the Mexican government has worked hard these past years to clean up any negative images of drug trafficking.

Further down the list, Central America saw a huge boost overall, especially in Costa Rica, Panama, and Nicaragua. The Balkans started to pick up quite significantly as well, with countries like Albania and Montenegro benefiting from the popularity of nearby Croatia.

On the other side of the coin, there are lots of cities that experienced low or negative growth that we would have expected to be higher. For example, Brazil overall didn't experience high growth rates before, during, nor after the World Cup and Olympics. This is something to keep an eye on as the World Cup takes place in Russia in 2018. New York also experienced slower than expected growth, possibly caused by continually high prices and a strong US dollar.

Not surprisingly, we also found a strong correlation with terrorism or political turmoil and a decline in traveler popularity. Turkey has seen a 38% decline in just one year as the country reels from numerous terrorist attacks and a failed coup. Similarly, the Egyptian tourism market has yet to recover to pre-Arab Spring levels, and Western Ukraine is only recently recovering from its Civil War, while Eastern Ukraine continues to decline. Just as Colombia’s market has grown with the improving political situation, Venezuela’s visitors have been cut by nearly 80% as the country struggles with unrest.

Like to travel? Help a startup out.

We just launched and would love to hear what you think

Your Instagram Feed Next Year: Predicting Future Up-and-Coming Destinations

Using our model of what makes a destination up-and-coming, we next used current growth trends to try and identify the most likely candidates for future growth.

We found that the destinations that currently exhibit the beginnings of major popularity are:

1. Canggu, Indonesia
2. Bocas del Toro, Panama
3. Sapa, Vietnam
4. Kampot, Cambodia
5. Montanita, Ecuador
6. Chiang Rai, Thailand
7. Cebu, Philippines

So don’t be surprised if pictures from these places begin to flood your Instagram feed starting in 2018.


There are a few caveats to this data that we should mention. The data comes from a single booking site, HostelWorld, so it is not capturing all backpacker traffic. However, in conversations with tourism providers, we understand that Hostelworld owns roughly 50% of the online hostel booking market, so we feel that this is a good proxy for the youth travel market.

Since our data comes from one platform, some of these numbers might be caused by HostelWorld itself growing as a company and capturing more of the market, rather than the overall market actually expanding. However, if HostelWorld does not have a significant presence in the market, it is a good sign that the market is not highly developed. Recent trends also show that the youth travel market has grown significantly over the timeframe of our data.

Another potential issue is that not all guests leave reviews. We do not know this true rate, and we are also making the assumption that all cities will receive reviews at a similar rate, and that this rate does not change over time.

Tell Us What You Think