As many have noted, it’s become crystal-clear that a business’ online reviews have been given exponentially more power and visibility with the new Google Places format.
And Google’s ramped-up focus on online reviews has seemingly just been further confirmed by a patent they were recently granted in November called “Systems and Methods for Reputation Management.” According to a recent, interesting wrap-up of the news (See: http://www.seobythesea.com/?p=4550)
…it seems Google is working on a way to devise an algorithm to, essentially, rate the raters – by ascribing a “Reputation Score” to online reviewers that could impact whether their reviews are included at all, along with assigning more or less “weight” to each review at Google Places.
The legalese/tech-ese of the granted patent is dense, and includes verbiage like this:
“A reputation management system…that assigns reputation scores to various types of entities including, but not limited to people, products, advertisers and merchants.
“…the reputation function is modified so as to remove portions of the function corresponding to nodes with negative reputations.”
“Upon convergence, reputation values for at least the reviewers and reviews corresponding to nodes that have not been removed from the reputation function are generated.”
As the blog author above points out, it remains unclear if this process of creating “Reputation Scores” applies only to reviewers who directly post at Google, or whether it would apply to reviewers at the sites Google Places pulls in.
But, as the blog notes, a few interesting concepts emerge:
*Reviews and ratings from reviewers that have “negative reputations” may not count in the final score for businesses, products, etc. they rate - and other reviewers they may score.
* If a reviewer only posts negative reviews, their reviews may be “removed” and not count.
* Reputation Scores aren’t just for reviewers, they’ll be assigned to individual reviews themselves.
It’s clear that Google is trying to hit upon an algorithmic solution that could address people trying to game the system - and one that would try to provide greater context for each review’s value. The patent suggests they’re reaching towards a pretty complex method: if a reviewer’s reviews are highly rated by others, then their Reputation Score goes up. But that Reputation Score, in turn, depends on the other Reputation Scores of people that rated that person’s reviews. And if a reviewer got bad ratings on their reviews by a rater that had a poor reputation, they wouldn’t be affected much, or at all. Etc., etc.
Google, of course, already lets people rate reviews, with this familiar question appearing under each review: “Was this review helpful? Yes - No - Flag as Inappropriate”
So, while this new patent activity strongly hints that Google is working on a solution that will provide a “score” to reviewers/reviews (based on some version of user feedback)…where it’s exactly headed, and how this would ultimately play out remains to be seen.
For instance…would the system essentially penalize people who don’t leave a high volume of reviews, because they won’t achieve a lot of positive ratings? (Should people who leave a lot of reviews be more highly rated/valued?)
And, even with some complex, situational algorithm - big questions remain about how any system would be able to filter out fake or “motivated” reviews if the reviewer is leaving positive reviews? I.e., unless a situation reached the levels of online consumer backlash that came crashing down on reviews posted by “Reputation Management” company, ReviewBoost, over the last few months? The outraged consumers who may have been wildly clicking “review not helpful” would, seemingly, in the Google equation, lower the value/score of those fake, positive reviews…
But how would an algorithm register/quantify their actual (enraged) comments, “This dealership SUCKS. All their reviews are fake…” Except as a “negative review”?
Online reviews have a huge impact on consumers, and the seriously increased visibility they’re given at the new Google Places just compounded that reality. This patent activity provides fresh evidence that the review/ratings component will keep evolving at Google, and continue to be a critical aspect of their local search model.
I think anyone would applaud ATTEMPTS to SOMEHOW provide better transparency to online reviews, and somehow zero in on all the inauthentic activity.
Fact: it didn’t take an algorithm for consumers reading fake reviews for dealerships posted by Review Boost to figure it out. It was absurdly simple: they clicked on the reviewers’ names and clearly saw that these “reviewers” left 5-star ratings for 10 dealerships in 10 different states - for a dozen hotels in a dozen different states - a bridal salon in California - a hair salon in Massachusetts – a skin diving company in Texas, and house cleaner in Utah – on the VERY SAME DAY!
(Question: why, after the online consumer uproar around the fake ReviewBoost postings, did Google then take down the user’s names, so consumers could no longer check what other reviews these people posted and figure out the “game”? Strange…)
Another fact: with whatever system is rolled out…with Online Reputation Management, an organically-driven, authentic process that generates believable, helpful, real reviews is inevitably going to be more important in the future.
The new Google patent makes one ponder in general: just what kind of system would work best to identify online review manipulation? What would represent the most meaningful way to ascribe a “Reputation Score” to reviewers? (I.e., somehow, when you think of new reviewer-rating systems to prevent gaming, new ways to manipulate the system do seem to multiply in one’s inventive mind...)