MicrostockGroup Sponsors


Author Topic: images progression follow up !  (Read 4419 times)

0 Members and 1 Guest are viewing this topic.

« on: September 23, 2017, 18:59 »
0
hi guys, i have started doing MS about one month ago, as per your advices here, I did pay attention to keywording ( as much as I can of course) and to assess how good my images are doing in the agencies search engines, I am randomly searching every few days for certain keywords ( which I know I have used) I mean specific keywords like Wales, certain type of flower macro, etc. so far, I came to a conclusion that some are climbing slowly through the search results ladder so I know they are keyworded more or less correctly and some are buried down the engine so I try to have another look into how I titled, described and keyworded them. the question here is, is this method theorically correct or are the agencies boost new stuff up the search? I am asking because I found out that some images for example have made it somewhere between the first and 5th pages of ( creative) in Alamy, and I am not sure if this is an indication that those images are doing well( even if there is no sale yet ) or is it just wrong to think this way ?!

Sent from my SM-G928C using Tapatalk



« Reply #1 on: September 23, 2017, 20:26 »
+1
Hi Wael

It is difficult to know how really Search Engines work as it code (of course) is secret. We are doomed to speculate and try to observe the behavior of sales as you are doing now.

It looks that sales influence images position in search, and maybe views... but the problem are these other "secret" factors.

Based in the fact that old popular images are not invincible, because some news images, even without sales are so close in search position that these old images, and also sometimes new images with few sales are on top... i can conclude that new images have some kind of preferencce and that when an image is getting old it start to lose popularity...

If the (teoretical) fact that old images lose popularity wasnt true, we would always see the same images at the top; and this is the reason why i believe that time online is one of these "secrets" factors. As image get old popularity is reduced.

Of course this interacts with other factors as sales and many other "secret" factors...

« Reply #2 on: September 23, 2017, 20:54 »
0
Hi Wael

It is difficult to know how really Search Engines work as it code (of course) is secret. We are doomed to speculate and try to observe the behavior of sales as you are doing now.

It looks that sales influence images position in search, and maybe views... but the problem are these other "secret" factors.


totally agree, there is certainly something i don't understand in this so-called" search engine algorithm ". quick example, i searched on Google for my flower macro shots by ( colourful flower close up ), found couple of my images in the first page; one from IS and the other from Alamy, then went to Alamy, repeated the search, found the image in the second page in their ( creative ) section, then went to IS , did it again, couldn't found the image until the 10th page of results. i find this really odd. it's either too complicated so no one can really get how it works, or it's too random so no one can really depend on something like this to have some sort of a " sales predictability "

« Reply #3 on: September 23, 2017, 21:16 »
0
never thought in my life i would be interested to read this  ;D
http://www.google.co.uk/patents/US8352465

« Reply #4 on: September 24, 2017, 03:57 »
0
The iStock search is not the same as Google's, so they will never display the same results. Every site has their own search and no one will behave the same.

You will have certain images that sell really well on some sites, but are never found on others.

« Reply #5 on: September 24, 2017, 04:33 »
0
Also bear in mind customers may get different results depending on their locations cookies  and (maybe) their customer profile maintained by the site.

Brasilnut

  • Author Brutally Honest Guide to Microstock & Blog

« Reply #6 on: September 24, 2017, 04:55 »
0
Quote
I am asking because I found out that some images for example have made it somewhere between the first and 5th pages of ( creative) in Alamy, and I am not sure if this is an indication that those images are doing well( even if there is no sale yet ) or is it just wrong to think this way ?!

You can check on Alamy measures to see which types of keywords clients are searching and how they correspond with your portfolio - views and zooms.

Even better is to see what "all of alamy" is searching on any given time period for up to a year and see if you can include some of these "hot" keywords within your images (don't do it if they don't fit the image as that's just spamming). I've attached a screenshot of the most popular keywords on Alamy for the past year.

« Reply #7 on: September 24, 2017, 06:18 »
0
Quote
I am asking because I found out that some images for example have made it somewhere between the first and 5th pages of ( creative) in Alamy, and I am not sure if this is an indication that those images are doing well( even if there is no sale yet ) or is it just wrong to think this way ?!

You can check on Alamy measures to see which types of keywords clients are searching and how they correspond with your portfolio - views and zooms.

Even better is to see what "all of alamy" is searching on any given time period for up to a year and see if you can include some of these "hot" keywords within your images (don't do it if they don't fit the image as that's just spamming). I've attached a screenshot of the most popular keywords on Alamy for the past year.
that's a very precious advice, thank you. unfortunately I don't have the time lately to shoot what is trending so I am just shooting what I usually shoot, that's why I am concerned more about how my images perform for a certain keyword and how that changes, trying to know if there are solid factors affecting that. it's a strange fact that most of my photos perform in alamy and Google search pretty much better than other agencies specific search.

Sent from my SM-G928C using Tapatalk


« Reply #8 on: September 24, 2017, 06:20 »
0


Also bear in mind customers may get different results depending on their locations cookies  and (maybe) their customer profile maintained by the site.

obviously I need to study more about how search engines work 😂😂

Sent from my SM-G928C using Tapatalk


« Reply #9 on: September 24, 2017, 06:23 »
0


You will have certain images that sell really well on some sites, but are never found on others.

any clue why this happens ? and how to meet every agency search criteria ?

Sent from my SM-G928C using Tapatalk


« Reply #10 on: September 24, 2017, 06:38 »
0


You will have certain images that sell really well on some sites, but are never found on others.

any clue why this happens ? and how to meet every agency search criteria ?

Sent from my SM-G928C using Tapatalk
Thats the $$$$ question some people think they know, personally I doubt it. One factor is customer base e.g Americans like shiny happy more  than us dour Europeans and some people just like the "look" of certain sites. Plain luck is also a factor ;-)

ShadySue

« Reply #11 on: September 24, 2017, 07:14 »
+1
The agencies order their searches as they choose, and can tweak them or radically change them any time they want.
This can be to your advantage or very much to your disadvantage, e.g. two of my best-selling images on suddenly stopped selling, one of which wasn't time sensitive, the other was very slowly (at least a decade) time sensitive, in content. This happened to each of them at different times, but both went more or less overnight from consistently top ten for a couple of years or more down to somewhere below 1000 when I lost the will to search further.

Alamy also changes from time to time. Don't necessarily believe what you read on their forum or even their official information about how search works. My own tests often prove that info wrong.
What Alamy mean by 'creative' is a mystery (but then so is 'creative' meaning 'can be used commericially' as e.g. iS use). I have one particular series on Alamy where they are all one class of subjects, but all totally different examples of that subject; all editorial. So they all have the 'group' name X, a lot of other keywords in common, but their specific subject a, b, c, d, etc. They all feature reasonably highly on a search for X (the 'class' of subjects) sorted by 'relevance', but some feature highly when X is searched on under 'creative' and some feature much lower down, and I can't see any connection between 'creative position' and 'files which have been zoomed or sold'.

Also, like iS, searching under 'new' in Alamy doesn't necessarily show the newest or most recently uploaded files.
There is a thing on Alamy whereby contributers use a 'nonsense' keyword (IIRC, 'ZRC') which won't spam any real search, but lets you know how each new tweak has moved your Rank - sometimes it can be several pages up or down. Apparently, Alamy has said that Rank doesn't feature so heavily in the search nowadays, but again, my (limited) tests would seem to prove that not fully true.

All you can do is keyword as accurately as possible. The rest is in the lap of the agency gods.
« Last Edit: September 24, 2017, 09:16 by ShadySue »

« Reply #12 on: September 25, 2017, 09:33 »
+1
Hi Wael

It is difficult to know how really Search Engines work as it code (of course) is secret. We are doomed to speculate and try to observe the behavior of sales as you are doing now.

It looks that sales influence images position in search, and maybe views... but the problem are these other "secret" factors.


totally agree, there is certainly something i don't understand in this so-called" search engine algorithm ". quick example, i searched on Google for my flower macro shots by ( colourful flower close up ), found couple of my images in the first page; one from IS and the other from Alamy, then went to Alamy, repeated the search, found the image in the second page in their ( creative ) section, then went to IS , did it again, couldn't found the image until the 10th page of results. i find this really odd. it's either too complicated so no one can really get how it works, or it's too random so no one can really depend on something like this to have some sort of a " sales predictability "

Of course as ShadySue says, agencies order their searches as they choose; they use different algorithm, and of course, they dont share it.

Anyway you have a big difference in algorithms when you compare whit google, and the difference is that google dont care the sales; it is not a factor for google image search algorithm, but it is key factor for MS search algorithm.

If you are really interested on it you must also know about a two very controversial theories in MS world.

First one is cap theory: in order to give a reasonable profit to "every" contributor your earnings are predetermined by agencies (secretly predetermined), like for example... you will eran about 700 US$ per month... and how well you do it will influence very little.

Based in this theory, and in order to work, search algorithm recalibrates the exposure of your images based on it.

Second one is adjustment of earnings according to your country: similar to the previous and based in the fact that 100 US$ per month is an amazing incoming in poorest countries but 1000 US$ per month is just a low earning in some first world countries, to keep contributors motivated, the profit is adjusted according to what is "a good profit" in your country.

Again search algorithm recalibrates the exposure of your images based on it (to give you as many sales to reach your predetermined income).

So, these theories are very speculative are are so controversial in this comunity, so i hope no one gets angry with me for considering and commenting on them.

P.S. some agencies (or at least one) recognized limiting the exposure of images based on giving a fairer profit to all contributors (as was commented by some forum user in previous post related to these theories).

« Reply #13 on: September 25, 2017, 12:48 »
+1
This is an extremely interesting topic, probably the most interesting to be discussed in a microstock forum, because if you understand how the algorithms work, you can organise your activity to maximise your efforts.
Strangely many people in this forum dismiss these discussions as conspiracies, as if agencies where totally neutral in the way they present their catalogs to customers.
By far the most important factor in the success of an agencies is how they propose their images to customers, in fact SS has a very sophisticated algorithm and is by far the best seller in microstock.

Personally I have detected some very evident patterns of behavior (especially in SS). Their priorities seem to be:
- Do not present to customers the same items over time (that is why SS switches between two different modes two-three times per month)
- Try to make as many contributors happy as possible
- Punish contributors that stop uploading
- Punish contributor that spam. I believe that the ratio files in portfolio/sales is extremely important, so people who upload tonnes of repeated material are lowered in rankings

« Reply #14 on: September 25, 2017, 15:02 »
0
This is an extremely interesting topic, probably the most interesting to be discussed in a microstock forum, because if you understand how the algorithms work, you can organise your activity to maximise your efforts.
Strangely many people in this forum dismiss these discussions as conspiracies, as if agencies where totally neutral in the way they present their catalogs to customers.
By far the most important factor in the success of an agencies is how they propose their images to customers, in fact SS has a very sophisticated algorithm and is by far the best seller in microstock.

Personally I have detected some very evident patterns of behavior (especially in SS). Their priorities seem to be:
- Do not present to customers the same items over time (that is why SS switches between two different modes two-three times per month)
- Try to make as many contributors happy as possible
- Punish contributors that stop uploading
- Punish contributor that spam. I believe that the ratio files in portfolio/sales is extremely important, so people who upload tonnes of repeated material are lowered in rankings

An amazing vision!

Thank you for sharing your knowledge Brightontl

« Reply #15 on: September 25, 2017, 15:53 »
+1
This is an extremely interesting topic, probably the most interesting to be discussed in a microstock forum, because if you understand how the algorithms work, you can organise your activity to maximise your efforts.
Strangely many people in this forum dismiss these discussions as conspiracies, as if agencies where totally neutral in the way they present their catalogs to customers.
By far the most important factor in the success of an agencies is how they propose their images to customers, in fact SS has a very sophisticated algorithm and is by far the best seller in microstock.

Personally I have detected some very evident patterns of behavior (especially in SS). Their priorities seem to be:
- Do not present to customers the same items over time (that is why SS switches between two different modes two-three times per month)
- Try to make as many contributors happy as possible
- Punish contributors that stop uploading
- Punish contributor that spam. I believe that the ratio files in portfolio/sales is extremely important, so people who upload tonnes of repeated material are lowered in rankings

That is very emotive and would be out of character for most businesses let alone algorithms. 
We see patterns in everything a strange human trait and how we try to make sense in the world.
The shutterstock search engine is in contant flux with numerous tests local and global going on tweaking parameters in conjunction with other parameters until it comes up with one which produces more sales, when it will be incorporated in to the main search. 
It is also a self learning beast with no emotions, it is neither happy or sad and cares less if the contributors are.  Why would it try to make a contributor happy if he is uploading crap?
The word punish may be substituted by reward as in if you are contributing more you would improve your search position, but, although I have heard Fotolia may have done this at one time, it is surely a consequence of several parameters like sales over time and new making someone who is uploading rise in the search and someone who isn't go down.  There is no need for a separate parameter for the number of uploads.
If they wanted to punish spam they wouldn't accept them in the first place.  I do agree with your last sentence, but it is not about punishing spam, but a consequence of a normal algorithm.

Try looking at with less emotion and more logic, I'm sure you will come up with something extremely interesting.

« Reply #16 on: September 25, 2017, 16:29 »
0
This is an extremely interesting topic, probably the most interesting to be discussed in a microstock forum, because if you understand how the algorithms work, you can organise your activity to maximise your efforts.
Strangely many people in this forum dismiss these discussions as conspiracies, as if agencies where totally neutral in the way they present their catalogs to customers.
By far the most important factor in the success of an agencies is how they propose their images to customers, in fact SS has a very sophisticated algorithm and is by far the best seller in microstock.

Personally I have detected some very evident patterns of behavior (especially in SS). Their priorities seem to be:
- Do not present to customers the same items over time (that is why SS switches between two different modes two-three times per month)
- Try to make as many contributors happy as possible
- Punish contributors that stop uploading
- Punish contributor that spam. I believe that the ratio files in portfolio/sales is extremely important, so people who upload tonnes of repeated material are lowered in rankings

That is very emotive and would be out of character for most businesses let alone algorithms. 
We see patterns in everything a strange human trait and how we try to make sense in the world.
The shutterstock search engine is in contant flux with numerous tests local and global going on tweaking parameters in conjunction with other parameters until it comes up with one which produces more sales, when it will be incorporated in to the main search. 
It is also a self learning beast with no emotions, it is neither happy or sad and cares less if the contributors are.  Why would it try to make a contributor happy if he is uploading crap?
The word punish may be substituted by reward as in if you are contributing more you would improve your search position, but, although I have heard Fotolia may have done this at one time, it is surely a consequence of several parameters like sales over time and new making someone who is uploading rise in the search and someone who isn't go down.  There is no need for a separate parameter for the number of uploads.
If they wanted to punish spam they wouldn't accept them in the first place.  I do agree with your last sentence, but it is not about punishing spam, but a consequence of a normal algorithm.

Try looking at with less emotion and more logic, I'm sure you will come up with something extremely interesting.

I think that Brightontl uses the term "punish" as an example of a factor that reduce popularity of an image, not literally.

Part of SS success is the big number of contributors which guarantee fresh content; so MS business is not just sell and sell without thinking in contributors, if a few of them monopolize first places at searches then there is no motivation for new ones, and this is not happends now, so an algorithm based in a balance between selling and maintaining the adherence of contributors is the most convenient in the long term.


« Reply #17 on: September 25, 2017, 17:18 »
0
This is an extremely interesting topic, probably the most interesting to be discussed in a microstock forum, because if you understand how the algorithms work, you can organise your activity to maximise your efforts.
Strangely many people in this forum dismiss these discussions as conspiracies, as if agencies where totally neutral in the way they present their catalogs to customers.
By far the most important factor in the success of an agencies is how they propose their images to customers, in fact SS has a very sophisticated algorithm and is by far the best seller in microstock.

Personally I have detected some very evident patterns of behavior (especially in SS). Their priorities seem to be:
- Do not present to customers the same items over time (that is why SS switches between two different modes two-three times per month)
- Try to make as many contributors happy as possible
- Punish contributors that stop uploading
- Punish contributor that spam. I believe that the ratio files in portfolio/sales is extremely important, so people who upload tonnes of repeated material are lowered in rankings

That is very emotive and would be out of character for most businesses let alone algorithms. 
We see patterns in everything a strange human trait and how we try to make sense in the world.
The shutterstock search engine is in contant flux with numerous tests local and global going on tweaking parameters in conjunction with other parameters until it comes up with one which produces more sales, when it will be incorporated in to the main search. 
It is also a self learning beast with no emotions, it is neither happy or sad and cares less if the contributors are.  Why would it try to make a contributor happy if he is uploading crap?
The word punish may be substituted by reward as in if you are contributing more you would improve your search position, but, although I have heard Fotolia may have done this at one time, it is surely a consequence of several parameters like sales over time and new making someone who is uploading rise in the search and someone who isn't go down.  There is no need for a separate parameter for the number of uploads.
If they wanted to punish spam they wouldn't accept them in the first place.  I do agree with your last sentence, but it is not about punishing spam, but a consequence of a normal algorithm.

Try looking at with less emotion and more logic, I'm sure you will come up with something extremely interesting.

I think that Brightontl uses the term "punish" as an example of a factor that reduce popularity of an image, not literally.

Part of SS success is the big number of contributors which guarantee fresh content; so MS business is not just sell and sell without thinking in contributors, if a few of them monopolize first places at searches then there is no motivation for new ones, and this is not happends now, so an algorithm based in a balance between selling and maintaining the adherence of contributors is the most convenient in the long term.

The big number of contributors is due to Shutterstock opening up entry to anyone who can produce one good image out of the 10 best images they can muster, no algorithm involved.

New images have always been weighted highly in the algorithm, though some would have it not as much as it used to be, but if profit has been weighted in conjuction with new then that would satisfy your assertion and confirm the conspiracy theory that 25cents an image is a lot more profitable than paying 38cents, but that in itself would not be a long term strategy, the exponential growth in content must slow eventually.

« Reply #18 on: September 26, 2017, 10:02 »
0
Quote
The big number of contributors is due to Shutterstock opening up entry to anyone who can produce one good image out of the 10 best images they can muster, no algorithm involved.

New images have always been weighted highly in the algorithm, though some would have it not as much as it used to be, but if profit has been weighted in conjuction with new then that would satisfy your assertion and confirm the conspiracy theory that 25cents an image is a lot more profitable than paying 38cents, but that in itself would not be a long term strategy, the exponential growth in content must slow eventually.

Everything is very speculative, I'm just saying it's logical to believe (at least to me) that the algorithm works in a way that allows new ones to have reasonable sales, not that new ones sell more than anyone else.

Last year SS changes approved rate to be contributor to 1 of 10 images, so they want new users, so they also needs a strategy yo keep those new users.

But basically the thing im saying is that SS limitate the exposure of users to give chance to other ones, if you have the most 1000 amazing cats photos, i think that your exposure is limitated to maybe 10 or 20 pics per page under "cat" search and so on.

« Reply #19 on: September 26, 2017, 10:33 »
0
This is an extremely interesting topic, probably the most interesting to be discussed in a microstock forum, because if you understand how the algorithms work, you can organise your activity to maximise your efforts.
Strangely many people in this forum dismiss these discussions as conspiracies, as if agencies where totally neutral in the way they present their catalogs to customers.
By far the most important factor in the success of an agencies is how they propose their images to customers, in fact SS has a very sophisticated algorithm and is by far the best seller in microstock.

Personally I have detected some very evident patterns of behavior (especially in SS). Their priorities seem to be:
- Do not present to customers the same items over time (that is why SS switches between two different modes two-three times per month)
- Try to make as many contributors happy as possible
- Punish contributors that stop uploading
- Punish contributor that spam. I believe that the ratio files in portfolio/sales is extremely important, so people who upload tonnes of repeated material are lowered in rankings

That is very emotive and would be out of character for most businesses let alone algorithms. 
We see patterns in everything a strange human trait and how we try to make sense in the world.
The shutterstock search engine is in contant flux with numerous tests local and global going on tweaking parameters in conjunction with other parameters until it comes up with one which produces more sales, when it will be incorporated in to the main search. 
It is also a self learning beast with no emotions, it is neither happy or sad and cares less if the contributors are.  Why would it try to make a contributor happy if he is uploading crap?
The word punish may be substituted by reward as in if you are contributing more you would improve your search position, but, although I have heard Fotolia may have done this at one time, it is surely a consequence of several parameters like sales over time and new making someone who is uploading rise in the search and someone who isn't go down.  There is no need for a separate parameter for the number of uploads.
If they wanted to punish spam they wouldn't accept them in the first place.  I do agree with your last sentence, but it is not about punishing spam, but a consequence of a normal algorithm.

Try looking at with less emotion and more logic, I'm sure you will come up with something extremely interesting.
Yes, I could have used reward constant uploading and reward a good "sales/number of images in portfolio" ratio, instead of using punish. I am not native English speaker.
I agree that SS search engine is probably some sort of AI, with learning capabilities, but still depending on what it is given as input. SS is certainly by far the most sophisticated engine of all agencies.

To me understanding how the algorithm work is the single most important thing in my work as contributor. All can can use for it is my own experiences with the patterns that I can see and reading what other contributor say.

I only talk about video sales, as I don't pay much attention to still images.
There is no doubt that there is some sort of "cap" in SS sales. I am not complaining at all about it, actually I think is a good thing. But this cap is dynamic and (in my experience) keeps going slowly up as long as the ranking go up.
Why they do that? because the want to keep a large number of contributor interested in uploading.

Regarding the fact of accepting everything. The main agencies for video now have practically a no-reviewer policy, and I think it works very well. Reviewers are expensive and are a total disaster, they have no idea of what could sell. It would be more effective to toss a coin instead of using reviewers. So the model for video now is to accept everything and then let the market decide (much much more effective).
If a contributor uploads stuff that don't sell (note that I am not using the term "bad" stuff), the famous ratio "sales/number of files" goes down and the contributor eventually disappears from the search engine. The same applies for a contributor that "spams", by uploading too many similar images


 

Sponsors

Microstock Poll Results