Why is it that these glitches ALWAYS affect only the contributers ?

Because it's DP we're talking about: they can't have buyers impacted, as there are none of them! :-p

Follow the money.

2. If you�re an established contributor, you make a higher royalty rate and SS keeps less of the money, so your work gets rejected more often and when accepted automatically gets pushed back in the search. That way new contributors can flood the site with images that get pushed to the top, and SS keeps more money by paying lower royalties.

I'm definitely not sure about that theory. I studied the numbers after reaching a new tier, and it did not have a significant impact.

being in a higher tier may mean you are more expensive, indeed, but it means your photos are more bankable, as it needs more than a few ducks in pound to reach the levels. Therefore, your return on investment may be higher.

Here are a few of my theories, after some disastrous months in terms of review:

-reviewers are still humans, but are blindly following the advice from an AI. The work they perform in terms of intellectual property detection or just people detection can't be conducted solely by an AI (they were for instance able to reject a picture that had a Louis Vuitton logo hidden in the background of a picture, half hidden by cables and branches, that did not appear in the keywords nor in the title). However, it seems they don't even consider to challenge the AIs quality analysis, explaining why we get bogus rejections.

-there are probably KPI and staffing issues with the reviewers. I am receiving bulk rejections for absurd reasons (for instance a batch of 10 files that will all have the same rejection reason, generally the title is not in English or the image is underexposed, while actually my batches are made of totally different pictures and topics). In the meantime, while commercial pictures are reviewed extremely fast, it can take up to several days for an editorial picture to be reviewed. It's still not 123RF, but it can be a pain in the ass while covering topics like the Coronavirus, that are really hot at the moment. My interpretation on this is that they are put under pressure to reduce the amount of approved files, while being understaffed. It can explain why they are actually not doing a real job at some time (like rejecting 10 pictures in a row based on a one second appreciation of one picture or approving in 30s pictures that don't have any interest but don't raise any red flag from their AI).

-when it comes to photos, their policies can be easily bypassed. As I previously wrote, I reopened a folder called "shutterstock rejects" where I copy all the rejected photos. On first submission, I can have a rejection rate of around 50%. I re-sbumit the next day the same files with no modifications, there are already chances for 50% of these previously rejected files to be accepted. Then, for the remaining ones, I will decrease the resolution while applying some light noise reduction and clarity adjustments and resubmit. Most of the files will be accepted at that point. For the last ones, I will continue progressively decreasing the resolution till it reaches the lowest resolution authorized. At that point, the initial batch got an acceptance rate of around 98%. I'm still losing a bit (up until now, I was around 99% acceptabce), and I'm wasting my time, but at least I am partially saving my ass.

-last thing: they are not very good at rejecting similar content. I am conducting a strategy to re-upload in commercial pictures that I had a long time ago uploaded in editorial. I discovered that, all over those years, my editing techniques have greatly improved, thus enabling me to better delete intellectual property elements, people, while providing a much better overall processing. I can therefore provide quality files on topics where there are currently a lot of demand (like Iran). Normally, such files should be flagged as similar. However, if you simply mirror the file, it become accepted. I did it as a standard on this category of pictures, at I have an acceptance rate of around 99%...

To summarize, my interpretation is that they are struggling to improve their review process while being cheap, which is doomed to fail. Their system is so cheap, however, that it is still possible to get around it. It's totally stupid, because it can force you to decrease the quality of files that were just fine, but did not fit their AI requirements in terms of focus and noise (especially in editorial), but it works.

SS has been very low to start the year.  However, this is typical.  Traditionally, for some unknown reason, sales are slow the first half of the month, but usually pick up towards the end of the month.  This week has been very good, though still not enough to make up for the lousy beginning of the month

I noticed the same. Over the past three to four years (before that time, the data isn't relevant for me), the days before the 8-10th of January are a disaster.

My blunt supposition is that there is a combination of the fact that many people are still on holidays, and the fact that there may be some yearly contracts (subscriptions or not) with SS that are not that easy to process (whether it's payments by the clients or just order management by SS), resulting in some delay, the time for the volume to be handled.

As some others were saying, it's hard to know given we don't have the X axis.

In any case, keep in mind that the main issue with Alamy comes from the comparatively low volume of sales. It's therefore hard to predict anything, as you would need a much bigger sample to make any reliable analysis. This explains as well why the sales volume may seem inconsistent on a 6 months or even a year analysis. I have made on Alamy in five years more money than the top 4 turds together, and yet, the sales volume in 5 years is inferior to the monthly volume of these 4 turds.

Whether it's RPD, RPI or just revenue, DP is the worst of the platforms I upload to.

After careful consideration, I finally decided to stay for another reason: personal advertisement.

I have noticed, indeed, that several of the buyers I had on DP were major news website from Balkans. Therefore, I am punctually using this as some references to my clients outside of the microstock world. It's indeed slightly less glorious than Alamy thanks to whom I can add references like BBC, Wired or The Independent, but on the local market, it's actually something pretty marketable.

That being said, one day or another, I think this will have to stop.

Well, my general principal, whether we're talking on stock or on other businesses (I have actually worked a lot for clients on paid ads) is about what you are selling. So, content is king, indeed, but it's not about an outstanding quality, but about niches.

The internet is big (not only in the microstock world) so you won't take it over. You need then to find niches that you can try to take over, whether it's location or topics/branches/businesses, check if there is a market for you, and take it over, and constantly experiment for new sectors. it's another principle linked to the long tail. In the end, it will be easier, and cheaper to control niches than trying to become visible on broader markets.

Actually, this graph is just basic economics, as a growth can't be exponential indefinitely, the data about the total payout, and the total revenue and benefits are what matter. About the last one, the overall evolution of the economic situation during the next months will be crucial.

That being said, as I was writing before, we can expect a few agencies to disappear and/or merged, as the market is saturated. TBH, I'm still surprised it did not happen before. I would definitely see Dreamstime being acquired, as well as 123RF disappearing a way or another (the same for DP). All of this, as well, could have some impact on the business and on Shutterstock.

Well, it depends on the type of microstock you're doing. I'm based in Southeastern Europe, mainly shooting editorial and travel.

The demand for southeastern Europe in these fields is actually incredibly low. So, shooting more or less the same topics, I get some significantly different performance.

In average, my RPI for a picture taken in the Balkans is almost three times lower than for a file taken in Germany, and between 5 to 6 times lower than for a Canadian picture. If I check my photos taken in Ukraine, it's even worse: the RPI is half the one of Balkans...

In the end, depending on what you shoot, there may still be a potential for you, even in a developed country, due to the much bigger demand.

Compared to last month, it's slightly better. If we take in account the fact that I got some refunds in August, the performance is more or less the same.

September (and october), however, are usually pretty bad months for me. If I compare the performance to last year, my revenue has been multiplied by 2.6. I checked a bit the reports, and I did not detect any outstanding sale during the month (the highest was around $4.

Therefore, I can't say this is an amazing month, but considering the season and the overall number, it is just fine.

That's actually surprising! While I noticed last month that most of my sales were coming from relatively old files (ID below 1.2 billion), the last week have seen a sharp increase of sales coming from files with an ID above 1.4 billion, plus an increase of the old and traditional best sellers.

In terms of quality, it is a bit strange, considering my files located around the ID 1.4bn are not as good as usual, due to the use of a lower quality lens during a trip, following some technical issues.

Furthermore, I have launched a search on a few terms that I'm used to search, and it seems indeed some newer files have been a bit pushed up.

That being said, considering the parameters taken in account in the algorithm, plenty of other factors could explain these differences...

So, to summarize, this month was a roller coaster.

The month started pretty horribly, before having around the 15th a lot of Single & Others, including a $50 one, which is pretty unsual (my portfolio is made of 99% photos), before slowly returning back to a normal pace.

In the end, I'm within my goals for this month. The $50 sale made it among my BME's. Deducting this sale, it would have still been an average month, slightly better than the previous ones. In terms of downloads, as well, it would have been in the average, even though the numbers were probably 5 to 10% lower than usual (the RPD was therefore higher, even without the $50 sale).

As I was previously saying, I have this conspiracy theory that SS is using sometimes the Single & Other column to renconciliate amounts after they screwed up something in the royalties. It's not the first time indeed that I'm seeing this scheme happening. For the moment, however, I would still say it's a kind of conspiracy theory, as it can be a coincidence, given that I lack of evidence, even statistically speaking.

In terms of algorithm, something happened, indeed. It seems the algorithm is promoting more intensively the older pictures. I got actually a lot of sales from files with an ID inferior to 1200000000, while I used to have in the previous months good performance from files between 1200000000 and 1460000000 (there were a few strategic mistakes that could explain why the most recent files are failing, even though it shouldn't be that radical). Even more, I got some pretty big surprises with 1st sales on files that were older (below the billion).

So, in the end, it's a bit of a stressful situation after a beginning of the year that was perfect. That being said, the algorithm changes, in the end, just transformed the way I'm selling, but not the average result. Now, let's see if the trend continues. The last years, October was a bit of a disappointment, we'll see if there's a seasonality here.