MicrostockGroup Sponsors


Author Topic: Shutterstock - Manipulating Search Algorithm Again?  (Read 3357 times)

0 Members and 1 Guest are viewing this topic.

« on: September 26, 2018, 08:14 »
0
Hello,

Something very fish is going on with Shutterstock.
I encoutered strange spikes, very highs and very lows.
For example:
 20-30 sales (various, new and old in different parts of the world).
Once the images in review got approved the next day 1 sale.

This happened for the 3rd time already.

30 sales, images reviewed, the next day 1 sale.

Today is the same scenario.

I really don't know what to make out of this but it seems that Shutterstock is struggling with their Search Algorithm and is deeply affecting my sales. There is no consistency anymore, just random sales (even sold few images from 8 years ago).

Anyone else seen this pattern?


Edit* : By "manipulating" I don't mean something "evil" going on. But so many strange changes in the Search Algorithm in such a short period of time makes everything hazardous.
« Last Edit: September 26, 2018, 08:20 by 2dvector »


nobody

« Reply #1 on: September 26, 2018, 08:16 »
+2
don't forget to add in the 'Cap' theory as well  8)

« Reply #2 on: September 26, 2018, 08:16 »
0
don't forget to add in the 'Cap' theory as well  8)


Sorry, not following :). You mean?

nobody

« Reply #3 on: September 26, 2018, 08:44 »
+4
don't forget to add in the 'Cap' theory as well  8)


Sorry, not following :). You mean?

you really are new  8)

« Reply #4 on: September 26, 2018, 13:11 »
+1
Hello,

Something very fish is going on with Shutterstock.
I encoutered strange spikes, very highs and very lows.
For example:
 20-30 sales (various, new and old in different parts of the world).
Once the images in review got approved the next day 1 sale.

This happened for the 3rd time already.

30 sales, images reviewed, the next day 1 sale.

Today is the same scenario.

I really don't know what to make out of this but it seems that Shutterstock is struggling with their Search Algorithm and is deeply affecting my sales. There is no consistency anymore, just random sales (even sold few images from 8 years ago).

Anyone else seen this pattern?


Edit* : By "manipulating" I don't mean something "evil" going on. But so many strange changes in the Search Algorithm in such a short period of time makes everything hazardous.

Don't waste your time trying to analyse sales. If you have 3000+ assets and a few years of data then you probably can make some sort of basic presumptions, otherwise forget it.

PZF

« Reply #5 on: September 27, 2018, 07:06 »
+2
Take no notice of Nobody. It's an 'in joke'. ::)
Yes, they are always tweaking the algorithm and really odd things do sometimes happen. Just how it is.
Be grateful you are getting some spikes. At the moment I seem invisible....!!!


Uncle Pete

  • Great Place by a Great Lake - My Home Port
« Reply #6 on: September 27, 2018, 11:40 »
+1
Take no notice of Nobody. It's an 'in joke'. ::)
Yes, they are always tweaking the algorithm and really odd things do sometimes happen. Just how it is.
Be grateful you are getting some spikes. At the moment I seem invisible....!!!

Part joke, part reality.  :) Yes the search is always changing.

Posted 11 September 2015 - 03:38 PM by vincent shutterstock

The search engine algorithm is proprietary, and there are numerous reasons why we are not going to share what it is. The algorithm is constantly tested and changed (several times a year, not just once every few years) with the simple goal of increasing downloads for all contributors.

Why and how we change the search engine algorithm is very simple: we test a new algorithm in a select market or on a random part of our customer base. If the new one gets more downloads we change it for all customers, if the new one does not result in more downloads we revert to the old one. We will always choose the algorithm that gets the most downloads.

Ultimately our goal is to keep it fresh and keep the customers interested (i.e. downloading) so an image showing on top of the search result for 8 years may be good for Laurin, but it is very bad for everyone else because customers do not want to see the same image every time for 8 years and will go elsewhere if it does as our tests have shown.

« Reply #7 on: September 27, 2018, 19:08 »
+4
Shuttestock does not make money by having buyers download more in a subscription based model. They tweak the algorthm in the hopes that buyes find what they want quickly without downloading unnecessary extra images. Lightboxes work against us as well. They can save dozens of images to show clients and only download the one they pick. Gone are the days of buyers downloading ten apples on white to let the client pick one.
There's no way algorthm changes are meant to benefit contributors.

« Reply #8 on: September 27, 2018, 19:45 »
0
Hello,

Something very fish is going on with Shutterstock.
I encoutered strange spikes, very highs and very lows.
For example:
 20-30 sales (various, new and old in different parts of the world).
Once the images in review got approved the next day 1 sale.

This happened for the 3rd time already.

30 sales, images reviewed, the next day 1 sale.

Today is the same scenario.

I really don't know what to make out of this but it seems that Shutterstock is struggling with their Search Algorithm and is deeply affecting my sales. There is no consistency anymore, just random sales (even sold few images from 8 years ago).

Anyone else seen this pattern?


Edit* : By "manipulating" I don't mean something "evil" going on. But so many strange changes in the Search Algorithm in such a short period of time makes everything hazardous.

Whatever "evil" is going on, please, please, please... bring it on!
I want more of those triple digit sales!

Uncle Pete

  • Great Place by a Great Lake - My Home Port
« Reply #9 on: September 28, 2018, 13:01 »
0
Shuttestock does not make money by having buyers download more in a subscription based model. They tweak the algorthm in the hopes that buyes find what they want quickly without downloading unnecessary extra images. Lightboxes work against us as well. They can save dozens of images to show clients and only download the one they pick. Gone are the days of buyers downloading ten apples on white to let the client pick one.
There's no way algorthm changes are meant to benefit contributors.

Nice summary, and the way the official line was spun into "increasing downloads for all contributors" is actually, increasing downloads for Shutterstock.

I never thought of light boxes anywhere, hurting sales. Interesting viewpoint. So do galleries hurt us too, because we make those to attract sales, when the same could eliminate sales? Somehow I think lightboxes might help sometimes, not just harm? But true the use could eliminate wasted downloads and decress our income, which would increase agency income.

I just question the part about buyers finding what they want quickly? How does accepting total crapstock, duplicates, images with inappropriate keywords and a search that isn't very adaptable for buyers to eliminate useless mismatched junk, helping make it easier to find what they want? Can you explain that to me?  :)

« Reply #10 on: September 29, 2018, 06:44 »
+2
Just opinions. And i can't be bothered or have the time to spend doing indepth analysis but i also don' t believe that sorting our potrfolios by popular or new is the same algorithm as the logged in buyers search algorithms?
But I am certain that nothing SS does is designed to make me or you money.

Noedelhap

  • www.colincramm.com

« Reply #11 on: September 29, 2018, 08:28 »
+2
Shutterstock's goal is to increase downloads and thus profits, and to stimulate you to upload more work; they don't care how many downloads you personally get. Whether it's 50% of the downloads for contributor A and 50% for contributor B, or 100% for A and 0% for B, the bottom line is all that counts.
That means the search algorithm is not designed to work against any specific contributor, it simply doesn't care as long as the bottom line is looking good.

« Reply #12 on: September 29, 2018, 08:53 »
0
Shuttestock does not make money by having buyers download more in a subscription based model. They tweak the algorthm in the hopes that buyes find what they want quickly without downloading unnecessary extra images. Lightboxes work against us as well. They can save dozens of images to show clients and only download the one they pick. Gone are the days of buyers downloading ten apples on white to let the client pick one.
There's no way algorthm changes are meant to benefit contributors.

Nice summary, and the way the official line was spun into "increasing downloads for all contributors" is actually, increasing downloads for Shutterstock.

I never thought of light boxes anywhere, hurting sales. Interesting viewpoint. So do galleries hurt us too, because we make those to attract sales, when the same could eliminate sales? Somehow I think lightboxes might help sometimes, not just harm? But true the use could eliminate wasted downloads and decress our income, which would increase agency income.

I just question the part about buyers finding what they want quickly? How does accepting total crapstock, duplicates, images with inappropriate keywords and a search that isn't very adaptable for buyers to eliminate useless mismatched junk, helping make it easier to find what they want? Can you explain that to me?  :)
I think they believe their own hype that their super duper intelligent search engine takes care of this.....they are miles from this and I think if someone comes up with a genuinely "intelligent"  search facility they will have a massive advantage.

« Reply #13 on: September 29, 2018, 08:54 »
0
Shutterstock's goal is to increase downloads and thus profits, and to stimulate you to upload more work; they don't care how many downloads you personally get. Whether it's 50% of the downloads for contributor A and 50% for contributor B, or 100% for A and 0% for B, the bottom line is all that counts.
That means the search algorithm is not designed to work against any specific contributor, it simply doesn't care as long as the bottom line is looking good.

In a suscription based model increasing downloads does not increase their profits.

« Reply #14 on: September 29, 2018, 09:00 »
0
But I am certain that nothing SS does is designed to make me or you money.

If we don't make any money then neither do they. So I can't really agree with that. I'm pretty sure they don't deliberately favour particular individuals and don't care if you or I make nothing at all, but they will care about the overall value of sales and will want to maximise the turnover.

« Reply #15 on: September 29, 2018, 09:02 »
0
Shutterstock's goal is to increase downloads and thus profits, and to stimulate you to upload more work; they don't care how many downloads you personally get. Whether it's 50% of the downloads for contributor A and 50% for contributor B, or 100% for A and 0% for B, the bottom line is all that counts.
That means the search algorithm is not designed to work against any specific contributor, it simply doesn't care as long as the bottom line is looking good.

In a suscription based model increasing downloads does not increase their profits.
Its only a partly subscription based model so unless they differentiate between buyer types thats only part of the picture.

« Reply #16 on: September 29, 2018, 09:03 »
0
Shutterstock's goal is to increase downloads and thus profits, and to stimulate you to upload more work; they don't care how many downloads you personally get. Whether it's 50% of the downloads for contributor A and 50% for contributor B, or 100% for A and 0% for B, the bottom line is all that counts.
That means the search algorithm is not designed to work against any specific contributor, it simply doesn't care as long as the bottom line is looking good.

In a suscription based model increasing downloads does not increase their profits.

Well, that's true, except that if buyers don't find what they want they will stop subscribing. Fewer subscriptions means lower profits. I doubt if they can do much to adjust the percentage take-up by subscribers of their download allowance.


« Reply #17 on: September 29, 2018, 09:05 »
0
Shutterstock's goal is to increase downloads and thus profits, and to stimulate you to upload more work; they don't care how many downloads you personally get. Whether it's 50% of the downloads for contributor A and 50% for contributor B, or 100% for A and 0% for B, the bottom line is all that counts.
That means the search algorithm is not designed to work against any specific contributor, it simply doesn't care as long as the bottom line is looking good.

In a suscription based model increasing downloads does not increase their profits.

It does at Istock for yearly packages and at shutterstock for smaller packages because we are on a percentage, so by your view of the algorithm we should be getting lots of little SODs, I don't see this, do you?

Noedelhap

  • www.colincramm.com

« Reply #18 on: September 29, 2018, 09:12 »
0
Shutterstock's goal is to increase downloads and thus profits, and to stimulate you to upload more work; they don't care how many downloads you personally get. Whether it's 50% of the downloads for contributor A and 50% for contributor B, or 100% for A and 0% for B, the bottom line is all that counts.
That means the search algorithm is not designed to work against any specific contributor, it simply doesn't care as long as the bottom line is looking good.

In a suscription based model increasing downloads does not increase their profits.


Increasing downloads, or subscribers*

« Reply #19 on: September 29, 2018, 12:15 »
+1
Shutterstock's goal is to increase downloads and thus profits, and to stimulate you to upload more work; they don't care how many downloads you personally get. Whether it's 50% of the downloads for contributor A and 50% for contributor B, or 100% for A and 0% for B, the bottom line is all that counts.
That means the search algorithm is not designed to work against any specific contributor, it simply doesn't care as long as the bottom line is looking good.
[/quote

In a suscription based model increasing downloads does not increase their profits.


Increasing downloads, or subscribers*

Increasing downloads is not the goal - increasing subscribers is where the money is.
And in theory increasing subscribers should mean more downloads.

nobody

« Reply #20 on: September 29, 2018, 15:59 »
0
The main goal or only main goal for SS is to increase their profits--- we serve their needs. Such is business...


 

Related Topics

  Subject / Started by Replies Last post
110 Replies
16158 Views
Last post April 11, 2014, 11:27
by tickstock
45 Replies
13007 Views
Last post May 05, 2016, 12:08
by Minsc
1 Replies
2159 Views
Last post August 16, 2017, 09:10
by YadaYadaYada
3 Replies
1459 Views
Last post August 17, 2017, 07:38
by JimP
34 Replies
7358 Views
Last post November 07, 2018, 18:41
by farbled

Sponsors

Microstock Poll Results