Gary, of View from the Wing fame, and Lucky/Ben of One Mile at a Time fame have both commented on their respective blogs about their (very similar) approaches when it comes to reading TripAdvisor reviews:
I use TripAdvisor for (1) the myriad of photos, a picture does speak 1000 words and (2) consistent comments across reviews — not normative judgments of good or bad, but frequently repeated data points. If reviewers frequently note mold in the bathtub, peeling wallpaper, stained carpets, that’s useful information. ~Gary Leff – from the On How to Parse Review Websites and Pick Hotels post
I don’t use TripAdvisor to read individual reviews and make decisions based on that, but rather look for consistent comments/feedback. ~Lucky – from the Tips for Becoming A Travel Blogger post
And according to an MIT study, which I stumbled across because Roosta posted a link to it in this exceptionally short thread on TravelBlog, it would appear both Gary and Lucky are taking the correct approach.
The Guardian wrote a piece about the study, and while the discussion hasn’t really taken off on the TravelBlog forums, The Guardian’s comments section is quite healthy.
Things with likes/upvotes tend to be seen by more people creating a snowball effect. For Example I often find myself only going through the highest rated and lowest rated comments on BBC articles then so they are more likely to be reinforced as good or bad. ~Mark Tomlinson
The herd effect – or a complex adaptive system to give it its mathematical name – is a well understood customer adoption model in offline and online marketing…? Maybe an interesting article, but not shocking news. ~dg000000
Is this article in response to one of my comments breaking the 200 recommends for the first time? Mind you, I didn’t think it was one of me best. ~vonZeppelin
Of course, these commenters are commenting on an article in The Guardian, so it is not a travel discussion per se.
How do you use TripAdvisor and other hotel review sites? Do you rely on the star ratings/scoring? Have you ever been burned by the “herd effect”?