A New Kind of Restaurant Review?

[johnb]

A while back I sent a PM to Don in which I suggested a new way of posting restaurant reviews that might be workable, and a way to broaden DR.com.  He recently got back to me, said he thought it contained some good ideas, and suggested I post my idea here, to see what others thought about it.

Here's the (slightly edited) text of the original PM.

============

Rocks:

Tim C's piece today about the problems of restaurant reviews and reviewers gave me an idea, and it seems to me DR.com might be the perfect way to go with it.

We all know the problems with Yelp, etc., and the limitations on professional reviewers. You've been trying but the DR review section still has limitations.

How about a hybrid approach, taking advantage of crowd sourcing but keeping out the riffraff, unqualified, and those with an ulterior motivation? Also making the reviews complete, easy to access, and current. Maybe this idea is already being done somewhere, but I'm not aware of it.

You could set up a new and different restaurant review section as an adjunct to the site with certain characteristics:

-It would be a wiki, so that there would be only one review of any given restaurant, not a string of posts to be piled through -- that would make it easy to use, more complete, and give it some authoritativeness. Also up-to-date.

-There would be limits on who would be allowed to post and edit the wiki reviews. For example, it would be limited to those who had accumulated a minimum number of posts on DR ("gold" members?), plus others who you would allow in such as industry folks, known credible individuals, whatever -- up to you. Presumably industry folks would have to declare affiliations and either be barred from commenting on those or have their comments in a separate section.

-Those who post/edit to it would have to use real names, which would be available to anyone who wants to look it up and see who said what and how the review has evolved over time (see below):

-Anyone could look through the editing history of any review and see the real name and changes edited in by any poster, from the most recent going back to the beginning). That would create a mechanism to prevent malicious edits and increase the credibility of the reviews.

-Anyone who has edited could also assign a numerical rating that would appear with the textual review, and these could be displayed and averaged.

In short, doing it this way would lead to worthwhile and reliable reviews, up to date, all accomplished by taking advantage of the crowd sourcing model.

And then you would become rich and famous just like those Zagat folks.

========

In short, what I proposed is a wiki restaurant review site (as an adjunct to DR.com), where each restaurant would have one entry only but the review would be constantly edited, corrected, and updated by a (slightly) controlled but potentially large group of voluntary reviewers.  Crowd sourced but reliable restaurant reviews. WikiRestaurants.

As in wiki sites generally, anyone would be able to review the edit history and see who has said what about the place, and what has been said in the past but changed or deleted by someone else.  This would help users form opinions about the review, i.e. the extent to which it is helpful and useful.

Thoughts?

[SeanMike]

I like this idea. Were you thinking of starting it with reviews already written on DR.com? Sort of having a "moderated" version of the board?

One advantage of this over a "traditional" review is that it'd be kept more up to date. Like when you go to a restaurant and it has a glowing Washington Post review...from 1999...on the wall.

As for who could post/edit/etc...I would definitely agree with them being authenticated and using their own names. I would also suggest that each user has a user page (sort of like you have here) that would list any industry affiliations, etc., that they might have such as where they worked in the industry, or if they're involved with ownership, for instance. Personally, I'd go as far as to say out of sheer comprehensiveness to list friendships - like how Don will mention when he's friends with a chef, etc. - but that might be a bit...weird...to work on.

I wouldn't make the limiting factor though number of posts. There are a lot of great members here on DR that don't post often, and there are people who post a lot who you might not trust with such a role. (I'm saying in a general fashion and not pointing at anybody here.) Perhaps if there was an editorial "board" who reviewed interested editors and approved them/rejected them/asked for more information...

[DaveO]

I think its a nifty idea.  I only wonder how much time it would take to manage it and to fight spam and purposeful inappropriate updates to the content.

[DonRocks]

I think its a nifty idea.  I only wonder how much time it would take to manage it and to fight spam and purposeful inappropriate updates to the content.

This is why "number of posts" is an important criterion (but not necessarily the only criterion). I'm interested in keeping this conversation rolling - I have absolutely no preconceived notions or ideas; John's PM to me sounded thoughtful, and I thought it was worth putting out there for public consumption. Keep talking - I'm listening.

[johnb]

I think its a nifty idea.  I only wonder how much time it would take to manage it and to fight spam and purposeful inappropriate updates to the content.

That's one positive aspect of the wiki model.  It is to a certain extent self-managing, especially if real identities are used, because if someone were to post something inappropriate, not only can somebody else immediately edit it right out, but that original post and subsequent edits are preserved for all time in the edit history page and linked pages, identified by who did it, and accessible to anyone.  Hence, there is a strong disincentive to post anything out of line in the first place, reducing the need for overt moderation.

Here is a sample of an edit history page.  This is from the Peter Chang entry in Wikipedia, which I myself have edited from time to time.  Takes a bit of plowing through, but the key point is that everything is there and/or can be linked from this page.