Maybe a shamless plug, but I work on mobile apps at Fodor's Travel. We've been doing print travel guides for 75 years. We now have apps, in addition to books.
We're professional curators. That means every restaurant review (and shops, hotels, etc) has been written by an actual human professional who's opinion we hope you trust (we're a little more upscale than Lonely Planet, but we have our hip moments). So in a way, we're the anti-UGC. When you have crowds of people, all the ratings become a noisey 3-4 stars. Including Burger King. That's why I think Fodor's content is a good alternative. We're in the travel space, not local, but, it's a refreshing alternative.
Places have good days and bad days (or even hours). Your reviewers could get made. While the reviews may be spot on, they are a one point in time random sample.
The other problem is that not everyone is the same. For example I absolutely hate "attentive" staff in a restaurant (let me eat in peace and if I want something I'll ask for it), but want the check as quickly as possible on finishing while others like to slowly wind down. A regular reviewer probably won't mention either of these because that is normal to them.
The existence of multiple reviewers means one will be more like me and would mention these issues which would make me go somewhere else while others who enjoy that kind of experience would be more likely to come. This doesn't make anyone right or wrong, but does show how different reviewers help and why the lowest review scores have the most informative reviews.
As a counterpoint, I don't like the same movies Roger Ebert likes. But his opinions are consistent, and I know well enough where we are different, that I can usually estimate from the text of his reviews how well I'd like a given movie.
Reviewers may prefer attentive staff, but the good reviewers (which exist) know that not everyone shares their preferences. That's the difference between a good review and a bad review: you can translate a good review into your own viewpoint and decide for yourself.
When I think of noisy one star reviews, I mostly think of the one star review for a microprocessor where the guy didn't use a heatsink and the motherboard melted. There are plenty of folk who come across well in writing but are monstrous in person, I'd never expect to trust their restaurant reviews, and I might not be able to identify them either.
The question is, are you going to find that one reviewer who is like you in the sea of 4-star-giving morons?
I love reading Ebert's movie reviews. I often disagree with him, but he is a mostly fixed point. I can tell if I'm going to like something because, over time, I have developed an understanding of where his tastes and mine align and where they differ. This kind of thing is a lot harder to do online—it requires real work and attention.
In a way, we're disagreeing about where we want to have our statistical anomalies. Do we want the reliable reviewer caught on a bad day, or perhaps recognized by the staff? Or do we want the sea of one-time reviewers, a few of whom had a statistically unlikely bad or amazing experience?
I just want to say, my mother has been using Fodor's Forums for years now to help her with planning her trips. So thank you for creating such a useful resource!
Huh, interesting, and I'm surprised I hadn't heard of Fodor's. I work at http://www.oyster.com/ -- we're kind of the anti-TripAdvisor as all hotels are visited and photographed by our pro photographers and reviewers. We've found the same thing with UGC in the hotel space. For the popular hotels on TripAdvisor, there are 1000s of reviews, and the aggregate rating is always 3-4 stars. In contrast, on Oyster.com there's one review written with the same guidelines as all the other hotel reviews (as well as 100s of high-quality, non-staged photos).
Man, I'm 30 and this makes me feel old. :) In the print travel guides market, it's one of the big names. It comes from the generation before the more backpacker-oriented guides (Lonely Planet, Rough Guides, etc.) and targets more of an upper-middle-class family vacation type of audience, but it's still fairly popular, within the print-travel-guides space.
It does seem that their move into digital guides is quite recent and incomplete, though.
Ya, I get that a lot.... I guess with less bookstores around, we have a lot less notoriety. You're right, high number of crowd ratings regress to some mediocre mean, less absolute dump 1 stars, less stellar 5 stars, i guess.
But not in the "back to square one" meaning, it is rather "next level of a spiral staircase". I believe there's still room for improvement (definitely there is over the '95 low tech and probably over what we have now) in the way we can combine internet and the '95 solution to find good <mover / restaurant / hairdresser ... etc>
Time will tell - lots of money goes into social, someone will do it right or at least we'll know what obstacles prevent this from happening.
This point doesn't apply very well to this case. The problem is that there was a lack of verifiable people on Yelp posting negative reviews. Yelp has many more semi-verifiable people on it (i. e. people who've left multiple reviews) than the typical person has friends with experience with a given rarely-used service.
If in this case the problem were likely to be solved by acquaintances, it would also be likely to be solved by Yelp, under its assumption that people need to leave multiple reviews to be considered credible.
No need for the 1995 solution, Yelp is just doing a very bad job at detecting bad reviews. It's usually rule #1 that in any type of "bad" user filtering that many people that are good users may not engage your site.
Q: Internet solution doesn't work. How do we fix it?
A: Use the 1995 solution but because it's not 1995 anymore you'll use Facebook and Twitter to ask your friends which is still using "the Internet".
Q: How can you verify the quality of a < mover / restaurant / hairdresser / babysitter / dentist / chiropractor / handyman / etc.>?
A: Ask your < friend / relative / neighbor / workmate / schoolmate / etc.>.
2009
Q: Same question.
A: Don't need no stinkin' people. We have the internets!
2012
Q: Internet solution doesn't work. How do we fix it?
A: Use the 1995 solution.