Evan Schuman's StorefrontBacktalk

Different media outlets have different policies regarding story comments. A strange incident this month here at StorefrontBacktalk illustrates the problems with bogus endorsements and how technology can sometimes help confront that issue—if someone is willing to spend the time investigating.

The story in question dealt with differences between online and in-store pricing.

We police comments on our stories very aggressively, to make sure that no SPAM gets through but also to guard against personal attacks, irrelevant comments (thoughts that don’t relate at all to the story being commented on) and—most frequently—vendors posting self-serving endorsements of their services disguised as a comment. For example: “This story is quite right that item-level RFID has a long way to go. Speaking of which, Smith RFID makes a darn nice package and it’s low-cost, too. We here at Smith have awesome customer service, as well.”

But the most problematic of such comments are the ones that try and fool readers. The Smith example just referenced—and, no, that’s not much of an exaggeration of the stuff some vendors try to pull—is at least honest, where the posters identify themselves as being with the vendor. We still screen out those comments because, in our opinion, that’s what readers want.

Many banned vendor comments, though, say nice things about a vendor without volunteering that the post is actually from within that vendor. Although this commenting approach does try to fool readers, the posters often at least use their real E-mail addresses (smith@vendor.com) or simply correctly note the company they’re with in the form they use to post.

This month’s strange post was different for a few reasons: It was made using a personal gmail account, included no reference to a specific company in the completed form and then made very nice comments about what turned out to be the poster’s employer. Indeed, the poster referred to that employer as “they,” which was quite clearly an attempt to make the endorsement sound legitimate and third-party-like.

But the truly odd part: The post said nice things about a retailer. (Note: It’s not odd to say nice things about a retailer. But it is odd for someone who works for a retailer—as opposed to a vendor—to bother posting bogus endorsements on a site mostly read by other retailers.

The phrasing sounded fishy, so we looked at the IP address. Sure enough, it traced back to the retailer being endorsed. Note to sneaky people trying to engage in comment SPAM: If you go to the trouble of using an anonymous gmail account and creating a false name, at least be smart enough to not use a computer on your corporate LAN or make the post while logged on the corporate VPN.

The poster in this case didn’t respond either to messages we sent or to those from that person’s own employer. But when we contacted the retailer’s IT department, they were extremely cooperative and the person was eventually tracked down and disciplined.

Given the action taken, we have chosen to not publicly identify the retailer involved. And that’s a shame, because its IT and corporate communications teams were wonderful and relentless in the hunt. We will say that it’s a midsize chain, so the exact time of the posting didn’t immediately reveal the poster. The company needed to run our IP address—coupled with the exact time—against all of its outgoing Web visits and then match that data against the people who were logged in at that moment and whether the connection was via VPN or attached to a particular internal workstation.

After much review, the culprit was identified this week. The access was from within the retailer’s offices. “The individual acted on his or her well-meaning but misplaced initiative,” said a communications director at the chain. “This individual did not receive any instruction or suggestion as to how to engage in this behavior, which we specifically address in our employee documentation and for which the individual has been admonished and counseled.”

But, the statement added, had the employee chosen to not use the corporate network (where he or she had logged in), that person would almost certainly not have been caught. For starters, had he or she not been linked to the chain’s IP address, we would have likely raised an eyebrow but couldn’t have concluded anything other than this person was an enthusiastic fan of the chain.

This incident is just one more reason to suspect the credibility of consumer comments, even when they are not billed as anonymous. Many services will help to answer E-mail questions and other programs will calculate the probability of a transaction being fraudulent.

But what about services/packages to identify and delete (or, better yet, prevent from ever being published) comments that are likely to be bogus? At the expense of not hearing from some enthusiastic but real fans, E-Commerce (and, soon enough, M-Commerce) would benefit from a higher quality of posts and, therefore, much more persuasive selling from legitimate supporters and critics.