Author Topic: Reshuffle on shutter ? (Read 14198 times)

Goofy · « **on:** January 30, 2014, 21:55 »

has anyone noticed a reshuffle on shutter? maybe I had too many beers and seeing things...

Beppe Grillo · « **Reply #1 on:** January 31, 2014, 02:30 »

What kind of reshuffle?

Ron · « **Reply #2 on:** January 31, 2014, 05:33 »

Happens all the times, your popular images constantly change, but the changes make no sense. My most popular image in DLs and $$ is dropping and images only sold once or never are close or higher. Doesnt make sense. Nothing I can do about it, and Jan 2014 was BME, so I guess its all good.

Shelma1 · « **Reply #3 on:** January 31, 2014, 07:20 »

They have an entirely new search result on mobile, at least. On my iPad I now see many more images, smaller and closer together, with continual scrolling and pages separated by a thin line. Looks more like google results. And things do seen reshuffled a bit. Plus, to switch from most popular (the default) you have to click on the small word "options" over to the top right.

Also, strongly vertical images are favored in the size they show up.

They must have been conducting a lot of behavioral research to make such a big change.

scottbraut · « **Reply #4 on:** January 31, 2014, 08:22 »

Hi all,

A quick note on how we do things: at any given time, we're running a number of tests. Those tests are typically targeted at a small percentage of the overall population. If a test wins (for example, if it drives more downloads and customers are more successful in finding what they're looking for), then it is released to the broader audience and the performance is monitored.

Tests can include changes to the user experience or the relevancy of search results. If you're seeing changes, it's possible that you're seeing a test that is targeted at a small audience, or that you're seeing the results of a test that won. These tests are thoughtfully conducted and changes are never rolled out broadly without careful analysis. For contributors, this might seem confusing in isolation, but it ultimately results in more successful customers, who then generate more downloads and more royalties.

Best,

Scott
VP of Content
Shutterstock

Ron · « **Reply #5 on:** January 31, 2014, 08:30 »

Quote from: scottbraut on January 31, 2014, 08:22

Hi all,

A quick note on how we do things: at any given time, we're running a number of tests. Those tests are typically targeted at a small percentage of the overall population. If a test wins (for example, if it drives more downloads and customers are more successful in finding what they're looking for), then it is released to the broader audience and the performance is monitored.

Tests can include changes to the user experience or the relevancy of search results. If you're seeing changes, it's possible that you're seeing a test that is targeted at a small audience, or that you're seeing the results of a test that won. These tests are thoughtfully conducted and changes are never rolled out broadly without careful analysis. For contributors, this might seem confusing in isolation, but it ultimately results in more successful customers, who then generate more downloads and more royalties.

Best,

Scott
VP of Content
Shutterstock

Scott, thank you.

Quick question, can you comment on why people see their personal most successful image drop in their portpolio popular sort and images that haven't sold or not sold nearly as much surpass the most successful image in ones portfolio.

Thanks again.

Shelma1 · « **Reply #6 on:** January 31, 2014, 09:08 »

Quote from: scottbraut on January 31, 2014, 08:22

Hi all,

A quick note on how we do things: at any given time, we're running a number of tests. Those tests are typically targeted at a small percentage of the overall population. If a test wins (for example, if it drives more downloads and customers are more successful in finding what they're looking for), then it is released to the broader audience and the performance is monitored.

Tests can include changes to the user experience or the relevancy of search results. If you're seeing changes, it's possible that you're seeing a test that is targeted at a small audience, or that you're seeing the results of a test that won. These tests are thoughtfully conducted and changes are never rolled out broadly without careful analysis. For contributors, this might seem confusing in isolation, but it ultimately results in more successful customers, who then generate more downloads and more royalties.

Best,

Scott
VP of Content
Shutterstock

Cool. I'm part of the test audience. That's a switch for me.

KimsCreativeHub · « **Reply #7 on:** January 31, 2014, 10:04 »

Interesting , thanks for the info Scott

My Very Best

KimsCreativeHub.com

Goofy · « **Reply #8 on:** January 31, 2014, 10:15 »

Thanks Scott!

ruxpriencdiam · « **Reply #10 on:** January 31, 2014, 11:30 »

Quote from: scottbraut on January 31, 2014, 10:39

Quote from: Ron on January 31, 2014, 08:30
Scott, thank you.

Quick question, can you comment on why people see their personal most successful image drop in their portpolio popular sort and images that haven't sold or not sold nearly as much surpass the most successful image in ones portfolio.

Thanks again.

Hi Ron,

It would be very difficult to answer something like this, because the algorithms are very complex; we run many different tests; and we see millions of searches. Anecdotal claims in the forums are very hard to track down - to understand what's going on, you would need to isolate all of the attributes of the image, the keyword that was searched, whether the image was in a test, etc...

As mentioned, the goal of all of this is to drive customer success, which ultimately delivers more downloads across the board and more royalties. It's easy (and understandable!) to worry about one image's specific search placement for a popular keyword search, but across all images and all searches, the net result is continuous improvement.

Best,

Scott

Nice to hear.

grey1 · « **Reply #11 on:** January 31, 2014, 13:05 »

Ha, ha! hilarious! whenever they announce a re-shuffle. Some will be poor, some will bust and some will get rich. What a joke to announce in a public forum. They should have known better.

heywoody · « **Reply #12 on:** January 31, 2014, 16:28 »

Make a real refreshing change from a site that rolls out half-baked idea after half-baked idea and does apply to the entire population.

disorderly · « **Reply #13 on:** January 31, 2014, 17:46 »

Quote from: heywoody on January 31, 2014, 16:28

Make a real refreshing change from a site that rolls out half-baked idea after half-baked idea and does apply to the entire population.

I agree. I like how judiciously Shutterstock management makes and rolls out decisions. It gives me confidence and makes me more willing to give them the benefit of the doubt when something doesn't sound entirely to my benefit.

sdeva · « **Reply #14 on:** February 01, 2014, 00:26 »

Thanks for sharing Scott, appreciate it

Whatever's underway worked well for me on the last day of Jan.

LesPalenik · « **Reply #15 on:** February 04, 2014, 06:00 »

Quote

Tests can include changes to the user experience or the relevancy of search results. If you're seeing changes, it's possible that you're seeing a test that is targeted at a small audience, or that you're seeing the results of a test that won. These tests are thoughtfully conducted and changes are never rolled out broadly without careful analysis. For contributors, this might seem confusing in isolation, but it ultimately results in more successful customers, who then generate more downloads and more royalties.

As mentioned many times before, the "Popular" designation has nothing to do with popular.
If you want to play with search algorithms, you can add another option to the available sort orders, and call it "flavor of the day" or as "recommended by chief cook" but calling it popular is not only incorrect, it is a deception and outright lie.

As contributors, we are getting shafted when our bestselling images are hidden and as buyers, we are misled and have to wade through numerous pages just to find something decent.

Furthermore it is naive to assume that if a small test yields certain results, the same results would be obtained across other portfolios if the same change is rolled out to other contributors. Each portfolio is different, each buyer wants something else, and whatever worked today may not work tomorrow. Calling the popular (or most downloaded) order popular would make the system simpler, consistent, and more honest.

grey1 · « **Reply #16 on:** February 04, 2014, 06:14 »

Yes Les! but I would strongly advice the agencies to just have three search alternatives and these would be the most honest towards us and the buyers.

1. wishful thinking
2. Scullduggery
3. not a hope in hell.

Now with these alternatives most here will know in advance they end up with a coronary occlusion or at best on the dole.

LesPalenik · « **Reply #17 on:** February 04, 2014, 07:37 »

Quote

Now with these alternatives most here will know in advance they end up with a coronary occlusion or at best on the dole.

Well, while some posters are destined to one of those two alternatives, I'm sure the others will pick up these concepts as inspirations for new shooting ideas.

grey1 · « **Reply #18 on:** February 04, 2014, 10:25 »

Quote from: LesPalenik on February 04, 2014, 07:37

Quote
Now with these alternatives most here will know in advance they end up with a coronary occlusion or at best on the dole.

Well, while some posters are destined to one of those two alternatives, I'm sure the others will pick up these concepts as inspirations for new shooting ideas.

I agree but even concepts do not grow on trees nowadays.

LesPalenik · « **Reply #19 on:** February 04, 2014, 10:54 »

Quote

The issue is the "popular" search. Not who is who. Thats a TV program. Anyhow Les is right. The popular, is not even remotely close to files being popular and how do they asses popular anyway? by downloads or views?

The Popular as implemented now has nothing to do with views or downloads. Well, maybe a little bit if these are new files.
If they are old files even with thousands of downloads, SS tries to push them out of sight. Maybe they are well meaning, but totally misguided and dishonest with their buyers. The designation as popular is misleading and deceptive. And very sad that they can't even find a proper name for it.

I can understand they they are trying to force the new files to the surface, but that order should never be called popular.
Popular means most downloaded.

grey1 · « **Reply #20 on:** February 04, 2014, 11:04 »

Quote from: LesPalenik on February 04, 2014, 10:54

Quote
The issue is the "popular" search. Not who is who. Thats a TV program. Anyhow Les is right. The popular, is not even remotely close to files being popular and how do they asses popular anyway? by downloads or views?

The Popular as implemented now has nothing to do with views or downloads. Well, maybe a little bit if these are new files.
If they are old files even with thousands of downloads, SS tries to push them out of sight. Maybe they are well meaning, but totally misguided and dishonest with their buyers. The designation as popular is misleading and deceptive. And very sad that they can't even find a proper name for it.

I can understand they they are trying to force the new files to the surface, but that order should never be called popular.
Popular means most downloaded.

BTW, take no notice of some geeks above. They have been like that for years. Yes thats what I thought, popular means in terms of downloads but then this so called popular search is all wrong.
In most searches, no way the files with most downloads are promoted, its really more like just a random search and because of spamming, thats the worst one.
However there is a small but valuable trick to avoid all this and pretty much make sure that your new uploads are in fact getting a good exposure.

Ron · « **Reply #21 on:** February 04, 2014, 11:05 »

Quote from: LesPalenik on February 04, 2014, 10:54

Quote
The issue is the "popular" search. Not who is who. Thats a TV program. Anyhow Les is right. The popular, is not even remotely close to files being popular and how do they asses popular anyway? by downloads or views?

The Popular as implemented now has nothing to do with views or downloads. Well, maybe a little bit if these are new files.
If they are old files even with thousands of downloads, SS tries to push them out of sight. Maybe they are well meaning, but totally misguided and dishonest with their buyers. The designation as popular is misleading and deceptive. And very sad that they can't even find a proper name for it.

I can understand they they are trying to force the new files to the surface, but that order should never be called popular.
Popular means most downloaded.

I have to agree with that. I can see how they add views to that equation, but most popular shouldnt be based on how quickly a file gets their first downloads or whatever fancy algorithm they come up with. My ultimate best seller is no longer my ultimate best selling image. Its dying, and SS killed it. Scott said, tests that generate the most downloads are pushed live, meaning more royalties, but that in my opinion is a partial fallacy. Yes it means more royalties, overall, not in my pocket. Unfortunately my bottom line is not important to SS test results. The SS RPD is though.

All they need to do is rename the tab most popular to Currently Hot and add another tab called Downloads, next to the existing Relevant and New tabs and its all sorted.

grey1 · « **Reply #22 on:** February 04, 2014, 11:30 »

Quote from: Ron on February 04, 2014, 11:05

Quote from: LesPalenik on February 04, 2014, 10:54
Quote
The issue is the "popular" search. Not who is who. Thats a TV program. Anyhow Les is right. The popular, is not even remotely close to files being popular and how do they asses popular anyway? by downloads or views?

The Popular as implemented now has nothing to do with views or downloads. Well, maybe a little bit if these are new files.
If they are old files even with thousands of downloads, SS tries to push them out of sight. Maybe they are well meaning, but totally misguided and dishonest with their buyers. The designation as popular is misleading and deceptive. And very sad that they can't even find a proper name for it.

I can understand they they are trying to force the new files to the surface, but that order should never be called popular.
Popular means most downloaded.

I have to agree with that. I can see how they add views to that equation, but most popular shouldnt be based on how quickly a file gets their first downloads or whatever fancy algorithm they come up with. My ultimate best seller is no longer my ultimate best selling image. Its dying, and SS killed it. Scott said, tests that generate the most downloads are pushed live, meaning more royalties, but that in my opinion is a partial fallacy. Yes it means more royalties, overall, not in my pocket. Unfortunately my bottom line is not important to SS test results. The SS RPD is though.

All they need to do is rename the tab most popular to Currently Hot and add another tab called Downloads, next to the existing Relevant and New tabs and its all sorted.

Quite right. It used to be based on views. If new file hardly get any exposure no matter how good they are. Never gets a download. Then what? just fades away and thats why uploading right now is quite simply a waste of time.

Shelma1 · « **Reply #23 on:** February 04, 2014, 11:34 »

Quote from: Ron on February 04, 2014, 11:05

Quote from: LesPalenik on February 04, 2014, 10:54
Quote
The issue is the "popular" search. Not who is who. Thats a TV program. Anyhow Les is right. The popular, is not even remotely close to files being popular and how do they asses popular anyway? by downloads or views?

The Popular as implemented now has nothing to do with views or downloads. Well, maybe a little bit if these are new files.
If they are old files even with thousands of downloads, SS tries to push them out of sight. Maybe they are well meaning, but totally misguided and dishonest with their buyers. The designation as popular is misleading and deceptive. And very sad that they can't even find a proper name for it.

I can understand they they are trying to force the new files to the surface, but that order should never be called popular.
Popular means most downloaded.

I have to agree with that. I can see how they add views to that equation, but most popular shouldnt be based on how quickly a file gets their first downloads or whatever fancy algorithm they come up with. My ultimate best seller is no longer my ultimate best selling image. Its dying, and SS killed it. Scott said, tests that generate the most downloads are pushed live, meaning more royalties, but that in my opinion is a partial fallacy. Yes it means more royalties, overall, not in my pocket. Unfortunately my bottom line is not important to SS test results. The SS RPD is though.

All they need to do is rename the tab most popular to Currently Hot and add another tab called Downloads, next to the existing Relevant and New tabs and its all sorted.

Well, now that I'm part of their test audience, I get to see their algorithm machinations in action. It's weird. On my laptop if I search a certain category where a couple of my images are on the rise, they appear near the top of the first page in results. On my iPad they appear on page 2 (which isn't really a second page any more, just a thin line separating it from "page 1").

Of course, I like the first algorithm better, but the folks with the "old" most popular files undoubtedly would prefer the second.

Shutterstock prefers whichever one brings them the most sales.

gbalex · « **Reply #24 on:** February 04, 2014, 11:47 »

Quote from: LesPalenik on February 04, 2014, 06:00

Quote
Tests can include changes to the user experience or the relevancy of search results. If you're seeing changes, it's possible that you're seeing a test that is targeted at a small audience, or that you're seeing the results of a test that won. These tests are thoughtfully conducted and changes are never rolled out broadly without careful analysis. For contributors, this might seem confusing in isolation, but it ultimately results in more successful customers, who then generate more downloads and more royalties.

As mentioned many times before, the "Popular" designation has nothing to do with popular.
If you want to play with search algorithms, you can add another option to the available sort orders, and call it "flavor of the day" or as "recommended by chief cook" but calling it popular is not only incorrect, it is a deception and outright lie.

As contributors, we are getting shafted when our bestselling images are hidden and as buyers, we are misled and have to wade through numerous pages just to find something decent.

Furthermore it is naive to assume that if a small test yields certain results, the same results would be obtained across other portfolios if the same change is rolled out to other contributors. Each portfolio is different, each buyer wants something else, and whatever worked today may not work tomorrow. Calling the popular (or most downloaded) order popular would make the system simpler, consistent, and more honest.

Well said, as a buyer on IS I found the content they served me infuriating and I left. Now we find shutterstock is playing the same games to increase revenue.

MicrostockGroup Sponsors

Author Topic: Reshuffle on shutter ? (Read 14198 times)

Goofy

Beppe Grillo

Ron

Shelma1

scottbraut

Ron

Shelma1

KimsCreativeHub

Goofy

scottbraut

ruxpriencdiam

grey1

heywoody

disorderly

sdeva

LesPalenik

grey1

LesPalenik

grey1

LesPalenik

grey1

Ron

grey1

Shelma1

gbalex

Related Topics

Sponsors

Microstock Poll Results

Sponsors