Their IPO states clearly that they prescreen images using proprietary inspection software. They still claim that every image still gets a set of eyes on it, but that does not jive with the whacky large batch rejections for out of focus when the images are tack sharp.  I do suspect that some images see real people and some don't, but I haven't a clue. But they do use pre-screening software to kick back images that do not meet the 1's and 0's requirements.

if you consider that most of the rejections are words identically , i would go with mantis to say machines.

They click a rejection buttons, of course the words are the same. Software prescreens for size, color, obvious gamut errors, and then it goes to a human. This prescreen avoids paid humans from wasting paid time. Then the flawed humans click random and wrong reasons because they don't want to get caught clicking for quick money, that can be done without actually reviewing. We send them back again and hope to get a serious review from somebody who cares.

The human review process is a bad joke, money for somebody in India who doesn't care, except to make quota. That's why the nonsense reviews and button clicks. Not robots.

The rejections at Shutterstock seems like random. They got the best sale volume but the worst review process, which clearly need to be fixed.

Their IPO states clearly that they prescreen images using proprietary inspection software. They still claim that every image still gets a set of eyes on it, but that does not jive with the whacky large batch rejections for out of focus when the images are tack sharp.  I do suspect that some images see real people and some don't, but I haven't a clue. But they do use pre-screening software to kick back images that do not meet the 1's and 0's requirements.

agree.

flawed humans click random and wrong reasons because they don't want to get caught clicking for quick money, that can be done without actually reviewing. We send them back again and hope to get a serious review from somebody who cares.

but that's just it... anyone with QC expertise will be able to instruct IT to produce a report to catch the reviewers who are not doing their job. you can tell by the persistent quick review per batch or the consistent OFF , bad lighting, poor composition.
unless the mid-mgt are the ones who also don't care and is just clocking in for quick money too.

india is this your assumption or is it true that it is being outsource to india. could be too, since india has a lot of ganja 

. . seriously, is it done by machine?

Yes, they are real people and they have Pizza Fridays, subsidized fitness plans and game rooms! http://www.shutterstock.com/jobs

Real people, i think that it is more easy to realize that when you are illustrator, reviewers indications are more clear, it cannot be a bot.

I'm sure any automated first-pass checks for focus, white balance etc would only be applied to photos.

750,000 + or - in a week. 225 reviewers. Do the math.

750,000 + or - in a week. 225 reviewers. Do the math.

i suspect bots because i had tested it several time last month with one or two tricky iffy stuff in a batch of perfect focus. i know a human's eye would not see it, but the bot would. and sure enough those exceptions got pulled out and rejected. also, perfect focus but tricky lighting like you would get
in a shot with specific point light with gobos, and true enough, they too got rejected as poor lighting or composition or out of focus. they were not oof nor were they poor lighting nor were they poorly composed.  a human reviewer would have approved it as all points were spot on.

so i am convinced bots are running the show. btw, what was that movie where the AI took over the city's traffic lights and tried to sabotage the govt???   one day ss will be taken over by this bot
...but then again, it won't matter because the ppl in there are already too stoned on marijuana

Not people they are ANGELS!