App Store review shenanigans, real and imagined
We first heard from iPhoneography and SCW last week about their two-man campaign to have skeevy developer Molinker pulled from the App Store (I included a link to their post, possibly a bit too subtly, in Sunday's post about the NYT story). Molinker was allegedly using the underhanded (and unfortunately under-caught) trick of distributing promo codes to 'puppet' reviewers who gave the company's photography apps five stars... and noticeably failed to review anything else on the store.
Since the independent reviewers of the same apps were uniformly one-star, this resulted in star distribution graphs for the apps that looked like sideways versions of the devil's horns. Satanic mischief, indeed. After the site delivered its investigation results via email to App Store top cop Phil Schiller, the offending applications (more than 1,000 of them) were summarily yanked from the store.
While it's good news for the App Store ecosystem that this kind of behavior is being monitored and corrected, and the iPhoneography team are to be commended for their diligence (although I doubt they'll be getting the requested "investigations reward for unearthing this blatant attempt at misleading and stealing from the public"), review manipulation schemes are seldom this blatant or easily identified -- and sometimes we start seeing them even when they may not really be there...
Our reader Noah emailed in this morning to strongly suggest that Gameloft was indulging in some chicanery around reviews for its game Modern Combat: Sandstorm, which has gotten a spike in one-star ratings due to a networking issue affecting multiplayer mode; users needed to adjust the UPnP setting on their wireless router to get the game to work as expected. On December 7, according to Noah's allegation, a slew of reviewers (many of whom have only ever reviewed one app, in similar fashion to the Molinker review squad) announced the UPnP fix while delivering a five-star review. Virtually all of the reviews suggesting the UPnP fix also have either eight or nine "found this helpful" votes... a bit of unexpected consistency.
Is this an example of App Store sockpuppetry, or just a legitimate audience response to unfair bad reviews for an otherwise good game? I'm generally inclined to give the developer the benefit of the doubt -- and, to simplify matters, I just asked them. Gameloft representative Sanette Chao, asked whether Gameloft had solicited or encouraged 5-star UPnP reviews, replied that "[t]his is a surprise to us, and we are not aware of anything of the sort."
The most suspicious bit, to my mind, is the 'helpful' votes; it really does appear a small group of people swept through the reviews looking for UPnP-related entries to vote up. Not that there's necessarily anything wrong with that, I guess; in this case, though, it's not the vendor doing the reviewing or voting.
Noah's reaction of suspicion, though apparently unwarranted, may be primarily a consequence of some of the genuine bad actors who are gaming the reviews system. The problem of false-fives isn't new on the App Store or in e-commerce at large; we've heard many, many stories of Amazon or iTunes reviews driven by friends, family or gift-carded strangers. The counterbalance to that is to encourage more legitimate user reviews to drown out the engineering; the only question is whether that's going to be enough to keep App Store shoppers' confidence high that the reviews mean something. Perhaps a review flag saying 'a promotional code was provided to this reviewer' would identify the worst of the abuses.
Photo from http://www.flickr.com/photos/webhostingreview/ / CC BY 2.0
