The mysteries of TripAdvisor

If I were a dishonest hotelier and asked my friends to write a positive review about my hotel, they would have a single place to write about. If I were to ask them instead to trash the opposition, they would have many more targets, and could write many more reviews. I might even think strategically enough to recommend that they didn't post the postive review of my place, and negative reviews of others under the same identity, to avoid making the commercial sabotage too blatent to my rivals...

"Now it could be that regular reviewers tend to get more choosy about handing out five stars. But it could also be that hotels get friends to plug their establishment and move it to the top of the rankings. So how is the traveller to tell?"

The trick is to read and judge from the quality of the negative reviews while read the positive ones with a pinch of salt. It's really the good negative reviews that will give you a more honest picture. Same logic shall be applied everywhere, not just TripAdvisor- caveat emptor.

"But it could also be that hotels get friends to plug their establishment and move it to the top of the rankings?"

Why even go through the trouble of getting third parties to shill their establishment rather than simply doing it themselves? I think though that everyone knows that Yelp, TripAdvisor and other such review sites are full of phony reviews, both positive and negative, and that there is a difference between using and relying upon information.

Yes, it's called "Astroturfing vs Crowdsourcing" and is a real problem in "Web 2.1"

That is why I don't bother with five-stars or one/no-star reviews. The middle ground tends to be more honest about an establishment.

The ranking of the place also comes as a very small factor when determining where to stay.

Sometimes I need to be close to the airport to catch an early-morning flight, or I want to be in the centre where all the nightclubs are so I can party it up, hence I don't mind the loud music pumping out into the wee hours - but not everybody has those needs (or wants) therefore, rankings are next to useless for me...

You should use Tripadvisor the way Ski Jumpers' forms are assessed:
Discard the top and bottom 20% of the reviews, and look at only the middle 60%.

Having said that, there was one example I experienced in Singapore where the radically split reviews were caused by the fact that the Hotel in question consisted of two distinct buildings, one of which was old (and got bad reviews) and the other still relatively new (and accounted for the good reviews).

Largely ignoring raves when a preponderance of them are from one-site reviewers is just one several ways I weigh TA reviews. What about the dates? Where are the reviewers from? (Some countries, naming no names, seem to produce really over-finicky travellers.) One even has to use the new keyword feature with care: looking for a NYC budget hotel recently I saw that "bedbugs" occurred several times, and I'd have completely rejected the place had I not noticed that the top review said something like "and thank goodness, no bedbugs for once!"

Negative reviews which are accompanied by a response from the hotel owner are the most interesting. Some hotel responses can reflect professional pride better than the shiniest compliment.

As something of a Tripadvisor addict, I do find a tendency for people to review places they feel strongly about--positive or negative--which may account for the clustering of reviews at the poles. Overall I've found the site very useful but not to be relied on as the sole source of information. And, for some reason, there are fewer transparently fake reviews on the French language site.

Out of curiosity - does Gulliver have an obligation to add a post about Trip Advisor on a weekly basis? Haven't we been over this territory a dozen times or more, just this year alone?

Trip advisor could simply put a guide based on their internal statistics. For instance, a rating of 1 to 5 stars regarding the statistical significance of the feedback (1 = no data or too little to form a valid opinion and 5 means there is a high number of contributors from different sources most of which have commented on multiple sights)

1. I saw a couple of one-star reviews of a hotel that my wife and I go to frequently. I wrote my first Trip Advisor review (5-stars). It was a Courtyard; I don't know the owner, since Marriott doesn't own most its hotels.
2. But how do you rate a Courtyard? Against similar hotels or on an absolute scale? Obviously, you don't expect the same experience as you would receive at a Ritz-Carlton or even a plain Marriott Hotel. On an absolute scale, you'd have to be a Ritz to get a 5-star review. No Courtyard would be worth more than 3 stars. I don't thing that Trip Advisor helps reviewers make that decision. Looking at the reviews, my feeling is that most reviewers base their ratings on expectations rather than absolute quality, but some go the other way. That's why you have to read the reviews rather than just look at the stars.

My actual experience in the hotels has been pretty consistent with what i find in TA
I only compare hotels that are within similar categories (4 stars against 4 stars, 200$ vs 200$ and so on and so forth);
I read negative and lukewarm reviews to check disastrous past events/consistent flaws in service/attitude;
I compare only the top box rating (excellent) among the hotels I'm comparing; i expect to find at least 70% of people who rate the hotel at the top bracket;
I disregard hotels with very few reviews (say fewer than 100).
It has worked well so far, with no exception (well, maybe one out of dozens).

but i myself have just reviewed 2 hotels so far, always to complain...

I've contributed 24 reviews to TripAdvisor over a 5 year period. 90% of my reviews are a favourable 4 or 5 star - why - because I do my research beforehand, using TripAdvisor and others - to check I'm not going to some hellhole.
When I do see an inconsistency in reviews (as per my most recent trip to Turkey), I take a look at the reviewers history. Not that they've done 10+ reviews, but for how long have they been contributing? Someone who's been a short-term contributor, and/or who only ever complains about the local food, not being able to get a full English breakfast, I tend to discount their contribution. Likewise the person who's only made 1 or 2 contributions, and is too effusive, giving 5 stars across the board, makes me equally suspicious - nobody's perfect.
I can still give a hotel an overall score of 4, but mark it down with a 2 or a 3 on some feature.

One of the ways I have learned to look out for false reviews is by checking the location where the reviewers are from. If the positive reviews are all from one location - or one country - then I would tend to be leery of the hotel. Especially with hotels in other countries, if there are reviews from tourists those tend to be more reliable than those of locals

I used TripAdvisor to plan a trip around Irish Bed & Breakfasts and the reviews were spot-on. Not that they were always 5-star, but one person's "they lacked hospitality!" is another person's "they left us alone to do our thing!"

I think much like Amazon reviews, it pays to be discerning. Extraneous details, common themes, story telling instead of just "this place is amazing!" lead to more-likely-to-be-reliable reviews, as does checking their reviews for places I have stayed to see if our tastes align. The internet is all about removing reliance on often-times coincidentally-acquired authority and moving towards critically consuming a breadth of content.

Sadly, as in all things online, TripAdvisor has changed radically in the last year. Once it was fairly reliable, now it's utterly unreliable, as much use as asking the man next to you on the train. Those who still occasionally have a look now call it TripConfuser. As in all these things, it's an aggregate of many sources that provides the only reliable guide. This is true of so many online guide sites especially restaurant sites. I have posted a negative comment on one, only to watch as, in real time, positive reviews were immediately posted to drive the negative one right down the list. After which there was no further activity. It is naive to assume that this was not done by the restaurant concerned - a famous one.

The is a current trend among marketeers and profile builders to saturate blogs and review sites with planned and targeted material, often called astroturfing.

In a minor way, hotels and other establishments can use multiple on-line personalities to seed blogs and ratings logs. In a major way, large corporations and governents can use it to mould public opinion.

There is even "persona management software" to automatically generate postings that appear to come from different individuals. The US Air Force has gone to tender for such software.

That TripAdvisor is being manipulated, probably by hotel groups and small boutique establishments is no great surprise and really quite small beer.