The Flaws In Tripadvisor Rankings

Categories Technology
20110718-221038.jpg

After over a month off from writing due to exams and holidays I am back.

It’s a common fact that many travellers rely on Tripadvisor.com as the Holy Grail for choosing hotels and attractions (e.g. restaurants). But are they informed properly? and do they always get the best value for money?

For those of you who have no idea what I’m talking about, TripAdvisor is a leading travel review site which has a primary aim of collating as many traveller reviews as possible for every hotel or attraction in the world. This is geared towards building a detailed picture as to which places give people the best value for money and which are places to avoid. An example of TripAdvisor’s use is from my own personal experience today. I’m currently staying in a coastal part of Los Angeles and it was my plan today to visit the famous Venice Beach. While researching Venice Beach in TripAdvisor it was evident the the area was now actually more crime ridden and dirty than picturesque beach paradise and so consequently the plans were adjusted. Many people use TripAdvisor in a similar way to the example mentioned and so it comes as no surprise that hotels and areas can be completely ruined by a few bad reviews, irrespective of whether they are merited. On the flip side, small establishments can achieve worldwide fame through good reviews which would have been impossible without these types of site.

While in general it could be argued that the reviews on tripadvisor are incredibly accurate there are numerous discrepancies that could be of note when searching for a good hotel. One of these is that tripadvisor is essentially a weighted survey due to the fact that the people who register and review will generally be of a similar demographic (frequent travellers, adults). It could also be infered that it is mainly people of a middle income/budget that will get any use from this sort of site due to the fact that people with low budgets have such a limited number of choices and people with high budgets won’t be as worried about value for money as more expensive hotels become ostentatious (more in demand as they become more expensive due to people trying to flaunt their wealth and gain exclusivity). This last factor means that the following diagram can be inferred.

In the diagram above the number of reviews peaks at an optimum price per night. While this may not seem like an issue it does actually have an effect if you dig deeper into how the tripadvisor hotels are ranked. The algorithm which ranks the hotels focuses primarily on the average star rating that the hotel has from reviews. If 2 hotels have the same rating (e.g. They both have five stars) then the decider is the number of reviews that the hotel has as this is taken as an indicator that more people prefer it. Now if you apply this rule alongside the approximate distribution of reviews it is clear that hotels of a medium price bracket are likely to be favoured over those of a higher or lower price in situations where the average rating is the same. If this sort of distribution was extrapolated further then there will be cases where some hotels have no reviews at all and hence won’t get a ranking.

It could also be said that the average member of TripAdvisor is likely to be a family traveller and thus will possibly not approve of hotels and resorts which have a greater focus on couples or young adults providing a more raucous atmosphere. In some cases these hotels may not merit a bad rating but due to them not being family friendly they may not end up with a true representative score. Additionally it is also unlikely that this majority demographic is going to provide numerous reviews for attractions such as nightclubs.

It would make more sense for TripAdvisor to try and rank hotels in a personalised way for each user so that the users personal requirements are taken into account rather than just generalising to every demographic. If a user is a young male looking for a lads weekend in Prague then the system should give more precedence to other user’s reviews who have done the same thing.

Another major problem that is widely documented is that of fake positive reviews. As mentioned earlier, a TripAdvisor ranking can make or break your hotel or attraction and so therefore an incentive is created for people to cheat the system. There have been numerous reported situations in which hotels have been found to be writing their own reviews in order to show themselves in a better light, it is easy to spot these types of reviews due to the fact that the PR representatives who right them don’t seem to be so savvy in covering themselves up. They will nearly always have generic user names (e.g. Iloveholidayinn1) and most of the time have opposite sentiments to the rest of the reviews. This sort of action can also work the other way through fake negative reviews whereby hotel owners or representatives write fake reviews in order to lower the ranking of a competing hotel. I feel that both of these cases are serious enough to warrant an investigation from the office of fair trading (or equivalent) under the accusation of false advertising and anti competitive behaviour.

To summarise, I wouldn’t say that you shouldn’t use TripAdvisor at all when booking a holiday because it is an incredibly helpful tool which can help you weed out some very dodgy hotels. However I wouldn’t use it as the sole evidence for choosing one hotel or attraction over another without checking other sites or hotel reviews from a similar demographic to yourself. For those of us who are of a more youthful, lively demographic who are less likely to want a family friendly holiday in Amsterdam I find that one of the best places to take a look at is the popular student website studentroom.com.

Currently studying Philosophy, Politics and Economics at St Annes College, Oxford University. I have a keen interest in applied economics, food and most types of sport.

Leave a Reply

Your email address will not be published. Required fields are marked *