Experiences as a stock image reviewer.
For seven years I reviewed images for Bigstockphoto, the online stock image agency till myself and 15 other reviewers were laid off in the summer of 2017. Not surprisingly, it was a result of an all to common occurrence in American employment—outsourcing to Asia.
What follows are my experiences as a stock image reviewer. All reviewers sign a Non Disclosure Agreement and the terms of my NDA have expired in September of this year. I will not list any names of supervisors or fellow reviewers. I will also not describe any proprietary processes, applications or user interfaces. I hope to give a good overview of what reviewing is like and what image reviewers experience.
Getting Started
Online stock agencies, call Microstock, got its launch in the early 2000’s. Based on the crowd sourcing model, it was a way for people sell images for under a dollar a picture. It remained controversial for many years as it was seemed to undermine traditional photography agencies. In many ways it has. Bad press implied the quality was poor, but Microstock has grown into a multi-million dollar business and many professionals sell their work thru these online sites. Just check the credits that following most TV shows and you will see such agencies as Shutterstock, iStockphoto and many others listed.
I had been into photography for 30 years both professionally and for fun before I started in 2005 with iStockphoto, Shutterstock and Fotolia. In 2009 I saw a job listing for reviewers at Shutterstock. I sent in a resume and forgot about it. A year later I got an email asking if I was still interested. I was even though I would be reviewing for Bigstockphoto, owned by Shutterstock. I went thru a brief training course that included a conference call with senior reviewers and the department head, followed up with sample images to review which were in turn, gone over by a senior reviewer. After that I was free to start reviewing and given access to the reviewer interface.
The Image Reviewers
Image reviewers hold one of the most important positions in a stock agency. Through the reviewers the content being submitted is filtered and available for sale. Few agencies I am aware of hire them as salaried employes. Virtually all are independent contractors working out of their homes and living all over the world. They exist under the crowd sourcing model and are paid less than the submitting image makers. However, unlike the image contributors, they do earn money on a much more consistent basis. No waiting around hoping an image might sell.
The reviewers are the most reviled people in online stock photography. Photographic forums often feature complaints for rejections, which by the contributor are always considered unjust and in personal experience, this is not without merit. Weird things can and do happen. A photographer once complained on a forum that his images of mannequins were rejected for not having model releases! I don’t know how this kind of thing happens or why a reviewer would make such a glaring and ridiculous mistake. One common complaint is for images rejected that are good sellers at other agencies. I have had the same thing happen to my images. This is what happens when a reviewer is granted complete independence. It is their call whether an image is commercial or not. For good or for ill.
Eventually, crazy theories start making rounds regarding what goes on in the reviewing process. Such as reviewers being bots and not real people. Let me assure you, there are real people reviewing stock. But theories like that only develop because some of the reviewing is bad or weird enough to make people think the image isn’t being actually seen by someone. Such as an AI bot coldly analyzing the image data. Not now, but perhaps someday in the future as AI grows in capabilities, it could happen.
In The Early Days
Basically, as reviewers, we were on our own. At the start it was apparently that not much had been worked out regarding many reviewing processes. A lot of things were done on the fly. That changed over time with lots of developments regarding admin access, communication with fellow reviewers and monthly webinars to learn the latest rules for accepting or rejecting images. Communication was simple at first just using IM. We could see our fellow reviewers but had no interaction with them. Towards the end of my time we migrated over to an IM app where all of the reviewers could chat and provide help to each other and get help and general group announcements from the senior reviewers. This greatly improved communication speed among all concerned.
One of the more important additions to the review process was a Wiki. As the image load increased, a flood of rules regarding copyright issues to adult content to all manor of issues, became a necessity. This was generally followed up by emails. At times it could be overwhelming with the flood of information to keep track of. One of the main concerns was copyright issues. This featured all kinds of things from color combos (John Deere), to unusual things such as the scraggly Cypress tree at the Pebble Beach golf course. (Who knew you could copyright a tree!) And it was more than just images. Well known brand names could not be used in titles, descriptions or keywords. What to accept and what to reject was a constantly changing process and the Wiki was a great help to keep track of the thousands of rules which would not be possible to remember on a day to day basis. Images that were rejectable could be redeemed later and be acceptable.
Basically, a reviewer was paid .06 an image. A bonus of $50 was awarded for meeting a quota of 10,000 images in a month. However, a common feature involving crowd sourcing is when they discover that the peasants are making too much money, they changed the quota to 7 days. It generally took me 2 weeks to make 10k so I never did get any bonus pay after that. Once I developed a knowledge of what to keep and what to reject, I was reviewing over 20,000 images a month. And that was working every day of the month to get there.
The Contributors
One surprise to me starting out is how many contributors were Russian or eastern European. Most model images I reviewed were of Russian, Ukrainian or anywhere in Europe. I never did understand why Russians took such great interest in super low paying stock photography, unless the exchange rate is very good for their currency. While very true, the slavs are a handsome people and make great models, but this results in a lack of ethnic diversity. Hence, most models were Caucasian with Asians and Hispanics being the second largest group and Black people trailed last. When I did see black models they were generally not Americans or from Africa, but European. In fact, I rarely saw models from America of any race.
Because of the large influx of images from Russia and surrounding countries, a huge problem arose with them not understanding English. This resulted in them not understanding the rules for submitting, keywording and describing images. Whatever translation apps they were using resulted in bizarre or hilarious descriptions for images. Here are some samples:
Merganser birds goo sander animals duck wild (Duck on the water.)
Various herbs are zerkleiert in mortar to prepare Globulis. (Various herbs with pestle and mortar.)
Corn heads of cabbage on an old table. (Ears of corn on a table. No cabbage in image.)
Little puppy cat climb on the plant. (Kitten climbing a tree.)
Corrugated wafer. (A waffle.)
As one can plainly see, these descriptions are obviously unusable and these images were declined. At times it was like reading descriptions of things from a member of a stone age tribe. The translation software must have improved over time as I saw fewer examples of this towards the end.
Since many contributors from eastern Europe cannot speak English, one mistake I keep seeing is their word cloud illustrations with multi-syllable phases chopped up. Such as the example below:
Guar
antee
It’s Guarantee. The separation is not at the syllable breaks so it appears as a two misspelled unpronounceable words. No hyphen used either, adding to this mess. At first I thought this was some language I was unfamiliar with. In that case, it would be rejected since Bigstock was English-only at the time. It was also the result of composition errors as this person had squared off the frame rather than leaving it rectangular. Regardless, this is unacceptable and was rejected.
VIP’s and Image Placement
For a while, some contributors were designated as VIPs. They were ranked 1, 2 and 3. Their images were to be given cursory review and accepted. In other words, they were given a pass. I assume this was done for photographers considered to be high-end and high production professionals. Their images were place high in the search routines. Some of the more well known photographers in Microstock were given this advantage. But over time, some VIPs (especially those in the 2 and 3 ranking) came to abuse their privileges with poor quality work, focus and noise issue, bad keywording, descriptions, grammar and spelling. After all, they knew whatever they uploaded would sail thru the review process. This did not go unnoticed and it was ended not long after I started.
At the beginning when reviewing images, we were allowed to leave messages for corrections to be made and then resubmitted. The previous reviewer’s comments was included with the image if the photographer had not followed instructions. Which was common. This most likely was due to language barriers. Though not a general rule, after three strikes I would decline.
At one point reviewers were allowed the option of placing a rating of 1, 2 or 3 on reviewed images for placement in the search. One mini-scandal occurred when it was discovered that reviewers that were submitting their own photos or illustrations to Bigstock would randomly come across their own submissions, pass them through and with a “1” rating. A group message via IM was sent out to knock it off. I know of no reviewer that was dismissed because of this. I never submitted to Bigstock to avoid an apparent conflict of interest. (Later this optional rating would reappear as five stars.)
I was often amazed at how people would not follow rules and guidelines. If they did they could have a better acceptance rate for their images. I saw a batch once from a guy that had a large portfolio of images submit a batch of model photos with no model releases. Maybe he just forgot that day. I had to reject them all.
And speaking of that, model releases were a huge problem. Quite often they were not filled out properly. Every kind of mistake they could make I’ve seen from no witness signature to missing names and dates. I saw an ID where the address on the ID did not match the address written for the model. One young photographer who was her own model, thought it might be okay to be the witness too. Nope, doesn’t work that way. Sometimes they would submit the model release upside down or sideways. All of these things are returns at the beginning and rejections towards the end of my tour of duty.
Keyword spamming was another issue. I found the worst offenders of this to be food photographers, who would often list items not seen in the image. If they had a picture of beef they would list pork as well. And vise-versa. Sometimes they would list a whole range of meats not seen—beef, pork, chicken, lamb. It was very rare to see a food photographer that did not do this. The issue occurred with seafood photos as well. If a photograph featured salmon, every fish in the sea got a listing (which could also include chicken, pork, etc.). It was not till much later that spamming, or wrong keywords for the subject, got to be a reason for rejection.
Image Content
Probably the worst thing about reviewing images is seeing the same old thing over and over again. Most contributors have no creative vision and are largely copycatting what they see from other photographers. So one sees a lot of flowers, cats napping on the sofa, brick walls, and the all time favorite, a cup of coffee imbedded in a pile of coffee beans. When they figure out how to photograph an object on a white background, out comes all the knickknacks off the shelf. Too bad that many don’t clean the gunk and dust off the things when viewed at 100%. After a while I thought, “These people would shoot a picture of a weed growing out of a crack in the street.” Sure enough, no long after, I saw it. The weed growing out of the crack.
After a while you realize most of these people don’t know what they are doing. Either with composing a photograph or understanding what market an image is for or how it is going to be used to communicate a message. This is after all, commercial photography and few to seem to understand the commercial needs of the marketplace. Some people were bound to take snap shots of whatever crossed their viewfinder. Ad agencies really don’t need pictures of dog poop or road kill covered in vultures.
The people I felt the worst for were not the rank amateurs. It was the talented people with professional equipment and photographic skills, excellent locales and professional models that were selling their pictures for pennies. I do not know it if was ever possible for them to earn a living this way. Yet they were abundant in their submissions.
Some photographers were quite good. One that stood out for me was a Frenchman that took pictures of a variety of subject matter, avoiding the clichés, including shots at crazy angles but well composed with good lighting. He has the talent of taking of ordinary things and making them look beautiful and interesting. For this article, I googled his work to see if he was still uploading. And yes, he is still shooting stock with the same high quality as the first time I saw it a few years ago. Jacques Palut. Go see for yourself.
I was once told by one of the senior reviewers that my acceptance rate just over 40%. He said their goal was to get an acceptance rate of 75%. I didn’t see how that was going to be possible. Many of these contributors were so incompetent that they could not get an image in focus even when using an autofocusing camera. Many images I saw were never post-processed when viewed at 100%. Many minor errors, such as sensor dust, could have been easily corrected. If Photoshop is too expensive there are plenty of lower cost imaged editors and some useful free apps available. It was as if they shot it and uploaded it straight from the camera. Experienced photographers know that the images out of the camera are just the first step in a proper workflow.
Many contributors were ignorant of many basic photography concepts but one I saw a lot regarded what a panorama format image is. It’s basically a long horizontal image with a 4:10 to 10:1 aspect ratio. Many people uploaded square or vertical images and called them panoramas. I did my best to correct this by sending them back with a note, but was eventually told to knock it off. As I was once told, the costumers will buy anything. This was evidences during performance reviews where I was sent pictures I had rejected and I was told I should have accepted them. Some were truly awful and I was annoyed that I should have accepted such rubbish. I got the feeling that they didn’t want the image quality of Bigstock to overshadow the quality at Shutterstock, the flagship agency and the owner of Bigstock.
Things I Never Imagined
I don’t know what I expected starting in as an image reviewer. I guess I was hoping for more professionalism from the contributors. Of that, I was sorely disappointed. A lot of people constantly tried to game system that has been set up for them to have an ease of use and to be able to market their images.
For example, a description for an image had a a 7 word minimum. Many sought to get around this by using periods, exclamation marks, dashes, spaces, or commas. I wondered how so many people learned to do that? I never saw this mentioned on any photography forums. Or why they thought using 7 words was so much trouble. Or why it wouldn’t be caught.
Other people liked to use their keywords and attempt to craft a description of an image from them. I guess that thought that would help advance their pictures higher in the search engine (which it did not). This only resulted in a grammatical meltdown. Example description for an image of a sandwich: Healthy wheat sandwich burger with BBQ grilled chicken steak cheese tomato rocket salad cucumber fried potato and mustard sauce served for eating on wooden board.
Yes, rocket was listed too!
Sometimes they would write a description that became philosophical or too opinionated. As in this description:
“A frozen leaf on the ground surrounded by grass and smaller leaves. The play on words was too tempting. I am dying, leaf me alone…”
One of the big surprises came from the guy in the Netherlands that found a way to bulk upload tens of thousands of images clogging up the image queue. I noticed it one morning when I was reviewing his CGI abstract watercolors. As I went thru the batches I noticed it was never ending. Picture after picture. I looked at the details of the upload and counted the pages. I don’t recall the exact amount but it was over 20,000 images. It was a shocker! I wondered if this guy had a team gathered to mass submit his files. I did not report this up the chain because I thought with their analytic tools they would have discovered this event. As it turned out, they did not know! It took them a few days to suss this out. The guy was using an app that encoded all the meta data into the images and this allowed him to flood the site. Anyway, he was asked not to do that again. They were very kind. I would suspended his uploading privileges for a month, if not forever.
I look back at something like that and consider how selfish and ruthless all of that was. This guy was not just gaming the system, he was abusing it. Besides that, it caused other contributors to wait even longer to get their images processed thru the queue. And all because one illustrator wanted to get the jump on the competition.
However, for the illustrators that played by the rules, it was not unusual to see portfolios in the 100,000 or 200,000 image range. I don’t know if these contributors had a group of people working for them or not. On the other hand, I saw portfolios where a photographer had a few hundred pictures accepted but thousands rejected. I don’t know what kept them going, but they kept banging away like John Henry with his nine pound hammer.
Regarding illustrations, the illustrators would often create logos. I always wondered about the use of logos since they could be sold multiple times. What if two different purchasers bought the same logo and tried to use it? What kind of legal battles would ensue? None that I know of.
One time, an illustrator was creating logos for hypothetical web sites. Curious, I decided to look up the URLs associated with the logos. It turned out some were legitimate web sites! With their own designed logos already in use. I sent this up the chain, and was told they appreciated my research but not to spend too much time on it. Okay, so I let it go. But I thought it was worthy of looking into. The designer had not bothered to do the proper URL research and was uploading graphics that if bought, would be useable.
Heading Home
Towards the end they decided to bring all of us reviewers on under the Shutterstock tent and we had to redo our application process which included watching instructional videos regarding image reviewing Shutterstock style, and as a final exam, review 60 image and score a 90% on the test. After that we were considered to be Shutterstock employees. Rather odd definition since we were still independent contractors with no benefits.
It started out being a free wheeling occupation and ended up highly restricted. Early on, I had my run of photos, editorial images, vectors or 3D illustrations. It was a nice variety of images and styles and help break up the monotony. Seeing loads of brick walls, pets, and the never ending parade of flowers can be mind numbing. Towards the end, new policies were in place to segregate the reviewers into review groups for these specific image types. I don’t recall why this was done, unless it was under the guise of efficiency. So I ended reviewed photographs and nothing else. We were divided up into Reviewer 1 and Reviewer 2 levels. The former restricted to 6,000 a week and the latter 12,000 images a week. That of course reduced the amount of income one could make reviewing. And that was probably the point.
When I was let go it followed no pattern. I had seniority over other reviewers but it didn’t matter. Different levels of experience both old and new were let go. I felt the worst for the newer reviewers as they had a crash course in learning the review processes just be tossed out. Really, why make us go through so much trouble with extra training just to dismiss 16 of us so quickly a few months later? Maybe it was a decision that was made in another part of the corporate structure and slowly made landfall with us. Or, could have been a last minute decision. Such is the nature of business in these times.
Let me state that while I enjoyed my time reviewing images I took no pleasure in rejecting people’s submissions. Having had my on images rejected, sometimes fairly, sometimes not, so I know the demoralizing feeling that rejection sends. But I learned from it and worked harder at improving my photography and illustration skills. I was hoping the rejected photographer would learn the same. Unfortunately, many I saw did not improve at all. They just kept hammering away, as if to show defiance, that you can’t hurt me.
The senior reviewers were great to work with and took care of us. If there was a payment error they took care of it quickly. If you can get a position like this, it makes for a great home job, especially if you have an ill family member to take care of, or as in my case, an elderly parent to look after.
