Online User Reviews: Can They Be Trusted?

Gaming the system involves more than just planting a review or two. It means creating multiple personalities and voices, crafting realistic conversations among those personalities, and using other tricks honed by stealth marketers and paid bloggers. (See Dan Tynan's "This Blog for Hire" for more information on how this works.)

"[These reviewers] are pretty fringe", says Ben McConnell, cofounder of the Church of the Customer Blog. "They come and go, change their names. They're like roaches, scuttling away when the lights are turned on."

Not surprisingly, the vendors, resellers, and meta-opinion sites we interviewed maintain that fake reviews are a very minor problem. There are safeguards to prevent most fake reviews from getting through, they say--and besides, readers aren't stupid.

"There's no way to vet the thousands of reviewers on Amazon," says Patty Smith, the company's director of corporate communications. "But we don't need to. When readers see 25 negative reviews and one glowing one--well, they can figure it out."

Screening Suspect Reviews

Still, most big sites have automatic and manual ways of screening out iffy reviews. Yelp's algorithms, for example, are on the lookout for suspicious patterns--such as a person who creates five new accounts and posts a positive review of the same restaurant from each account. Suspect reviews can be automatically suppressed, the reviewer's IP address blocked, and so on. Yelp caused a ruckus this summer when it yanked a number of reviews. At the time, Yelp said it was disturbed about positive reviews that some business owners had swapped with other owners.

Epinions has a panel of 10 to 15 readers in the topic area that screens reviews before they're published. "These core members will pelt the reviewer with questions and ask them to flesh out the review," says Alisa Weiner, Epinions' vice president of online comparison shopping. "They have a very high bar for what they consider a useful review."

The real secret weapon many sites rely on? Community policing. Namely, dedicated (dare we say obsessed?) readers who whomp on any review that's suspicious. "It's hard to game our core base of reviewers--they're very protective," says Jeremy Stoppelman, Yelp's CEO and founder. But he admits policing can sometimes turn vigilante.

If a dubious review appears--say, a glowing review for a known lousy restaurant--Yelpers will sometimes pile on with one-star reviews. "And of course," sighs Stoppelman, "we have to wade in and undo the mess."

Beyond community policing, what about real policing? "Fake reader reviews would violate section 255.5 of the FTC guidelines on the use of endorsements and testimonials in advertising," says Frank Dorman of the Federal Trade Commission. Such deceptive practices could conceivably land the "reviewer" in hot water. But none of the sites we talked to, from Amazon to Yelp, have ever pursued such legal action. Self-policing and filftering technology, in their view, are more than enough.

What's a consumer to do? Here's some advice from online shopping pros, flinty-eyed reviewers, and others in the know.

  • 1
  • |
  • 2
  • |
  • 3
Join the Discussion
blog comments powered by Disqus
You Might Also Like...