4 ms·
mass deletion of negative reviews is automatic and has been done before if it looks like spam / brigading (which this probably looks like). i think it's a reaso
by kbar13 6y ago
mass deletion of negative reviews is automatic and has been done before if it looks like spam / brigading (which this probably looks like). i think it's a reasonable explanation.
https://android-developers.googleblog.com/2018/12/in-reviews-we-trust-making-google-play.html https://android-developers.googleblog.com/2018/12/in-reviews...
https://9to5google.com/2018/12/17/play-store-anti-spam-reviews/ https://9to5google.com/2018/12/17/play-store-anti-spam-revie...
- sneak 6y agoI don't think the negative reviews of Robinhood in the last day qualify as "spam" under any definition with which I am familiar. I also see nothing wrong with brigading, if you are an actual legitimate user of an app and are unhappy with it. Businesses solicit positive reviews all the time.
- tptacek 6y agoEvery notable online review venue works to resist canvassed reviews, which break the whole concept of online reviews.
- sokoloff 6y agoAlmost every free iPhone app I install has a "enjoying our app? Rate it in the app store!" interstitial at some point in the user journey. Is that "working to resist canvassed reviews"? (I agree they should; I don't agree they are resisting it.)
- tptacek 6y agoIt is not my claim that any online store does an effective job of policing canvassing, only that they all attempt to do so. No crowdsourced online review source is trustworthy. But they're untrustworthy because of canvassing. I'll not also that soliciting a review of an app is not quite the same thing as organizing an effort to get a huge directional shift in reviews.
- sokoloff 6y agoAgreed. Both are pretty clearly canvassing though, in my estimation.
- spunker540 6y agoIt’d be neat if rather than just display a lifetime average they displayed a stock ticker for the ratings. You could see the app was good but then very lowly rated for a few days before ticking back up for example
- FireBeyond 6y agoMaybe more tangential, but isn't that the app itself asking, like a gatekeeper? I think if Apple generates that message or its triggered, then it's one-shot - if the user says "No", it will never be seen again (to be clear, to me and most, this is desired behavior). So instead you have the app asking things like this, or "can we send you notifications?", and so if you say no, it can ask later, and not trigger the OS interaction until just-in-time.
- kbar13 6y agoyeah i'm not saying robinhood is in the right here, i personally. i just think that this may not be an example of apple/google conspiring to take action because of any triggers from the past two weeks' events.