MicrostockGroup Sponsors


Author Topic: New best match penalizes files with high downloads to views?  (Read 9747 times)

0 Members and 1 Guest are viewing this topic.

PaulieWalnuts

  • On the Wrong Side of the Business
« on: November 10, 2008, 08:52 »
0
Just noticed a pattern.

Several of my better selling files literally went from the front of my portfolio to the last page.

At initial glance any files that have a view/download ratio better than 10:1 get penalized. Some examples of files that have been moved to dead last:

- 143 views / 22 downloads
- 5375 views / 777 downloads
- 165 views / 23 downloads
- 348 views /51 downloads

And now the files that are still at the front of my portfolio:

- 542 views / 24 downloads
- 359 views / 20 downloads
- 608 views / 29 downloads
- 729 views / 45 downloads

Seems to give preference to files with a 10:1 - 20:1 ratio. Anything better gets heavily penalized. Anything worse just doesn't get preference.

Why the heck would they penalize files that get high downloads to views? Isn't a high view/dl ratio an indication that buyers feel it's a very good image?

Anybody else seeing this trend in their portfolios?


« Reply #1 on: November 10, 2008, 09:22 »
0
my best file has a 7:1 view to dl ratio and got moved from 1st to page 22 of my portfolio. Obviously you can imagine what happened DL wise. Why they - the idiot powers that are - do this is beyond my comprehension? The only thing saving my performance is the high diversity of my portfolio. I got no signature style.

« Reply #2 on: November 10, 2008, 09:33 »
0
The LAST PICTURE of my portfolio has 16 DLs and 118 views. QED. Of course now it will neither be seen nor downloaded. These f****** idiots at IS. Note to self: Do not upload highly specialized pictures that target a tiny market because if they fit too well they get KILLED. Morons.

« Reply #3 on: November 10, 2008, 09:35 »
0
I too think DL /views ratio might have  some sort of effect on best match and   this is why I got suspicious if someone trying to increase the view of others files to manipulate the search results.
best match mystery goes on...

« Reply #4 on: November 10, 2008, 10:13 »
0
The LAST PICTURE of my portfolio has 16 DLs and 118 views. QED. Of course now it will neither be seen nor downloaded. These f****** idiots at IS. Note to self: Do not upload highly specialized pictures that target a tiny market because if they fit too well they get KILLED. Morons.

Maybe you should extend your opprobrium to those who were gaming the system this time last year with their buying gangs shooting a selection of seasonal images right to the top of the best match - self perpetuating incredible dl/view ratios. No doubt this best match is over reactive to that scenario at this peak buying time but moronic behaviour is not the sole prerogative of istock.

« Reply #5 on: November 10, 2008, 10:14 »
0
I just looked at the last page of my portfolio and there are several of my best sellers.  This best match is crazy and the only good news is they should be changing it this month.  Perhaps we should tell the buyers that the good stuff is on the last page :)

« Reply #6 on: November 10, 2008, 10:39 »
0
The LAST PICTURE of my portfolio has 16 DLs and 118 views. QED. Of course now it will neither be seen nor downloaded. These f****** idiots at IS. Note to self: Do not upload highly specialized pictures that target a tiny market because if they fit too well they get KILLED. Morons.

... opprobrium  ...

I get the meaning, but I still have a hard time using this word properly in a sentence even after wikiing it. I can only imagine how much the IS vocabulary likes your keywords - if they understand them. Thanks

« Reply #7 on: November 10, 2008, 10:57 »
0
Another possible factor behind this observation could be a self limiting brake on runaway files.

I'm sure we've all scratched our heads wondering why one generic brickwall/flower/handshake has sold a thousand in six months. Now if that file suddenly dies the owner will notice a great deal more than the possible hundreds of beneficiaries of one extra sale here and there. But in the long run a smoothing out of these hyper files would benefit many.

Obviously this would also possibly disrupt successful unique files, but they ten to have lots of 'just looking' views too.



« Reply #8 on: November 10, 2008, 11:58 »
0
Another possible factor behind this observation could be a self limiting brake on runaway files.

I'm sure we've all scratched our heads wondering why one generic brickwall/flower/handshake has sold a thousand in six months. Now if that file suddenly dies the owner will notice a great deal more than the possible hundreds of beneficiaries of one extra sale here and there. But in the long run a smoothing out of these hyper files would benefit many.

Obviously this would also possibly disrupt successful unique files, but they ten to have lots of 'just looking' views too.




Exactly. Those of us who aren't exclusive know that in general, best sellers on IS are nothing special, or they'd be best sellers on every site.
I for one would much rather see lower sales on a larger number of files. That would give contributors a better idea of what is useful to buyers.
Also, runaway best sellers encourage copying.
I quite like the new best match, even though it has affected me badly. I can see that I've got more sellable files than I thought, and that's helpful:-)


« Reply #9 on: November 10, 2008, 12:11 »
0
I think it also would be helpful to have similar files in your portfolio - that could compensate for the hit that the popular file gets. Diversification is the name of the game at IS.

lagereek

« Reply #10 on: November 10, 2008, 12:14 »
0
Theres only one consolation: Thank GOD we never became Exclusives at IS.
At the same time: pitty the ones that did.

                         

« Reply #11 on: November 10, 2008, 15:11 »
0
Let me see if I did it right. I searched for my name with sort by Best match.  If this is right, then my results are quite varied.

The six first images get: 25-27-49-33-26-72  (ave:39)

Six images in the middle of my list get: 23-37-55-49-60-46 (ave:45)

I can't use the last page because I have many zero downloads...

All my best sellers are in the first 4 of 12 pages, so I don't know if I did it right or if my reuslts are simply not implying a heavy weight for views/dld.

Regards,
Adelaide

« Reply #12 on: November 10, 2008, 17:07 »
0
The easiest is to click on your name, than on the portfolio link on the right and make sure Best match sorting is selected. The question is: Do you have any images that have less than 10views/dl. If not do not worry about it because it just means that your images are flying below the penalty threshold.

« Reply #13 on: November 10, 2008, 17:37 »
0
The easiest is to click on your name, than on the portfolio link on the right and make sure Best match sorting is selected. The question is: Do you have any images that have less than 10views/dl. If not do not worry about it because it just means that your images are flying below the penalty threshold.

Sorry, I don't get it...

How could your portfolio be sorted according to a Best Match; best matching what? There must be a reason why my photos shows in a given order when I look at my portfolio, but I think it has nothing to do with a search with keywords.

Am I wrong?

Claude

« Reply #14 on: November 10, 2008, 17:43 »
0
I don't have many, but I have some falling in the range between 8-12 views/dld.  Most of them are also my best dld/mo, so I can't tell if the latter is weighing more.

The best in V/D is not in my first page of BMs.  Yet the 6th in my best match is one with 72 V/D and that haven't sold for two years (!!), so it's not to say that images with many uploads recently also got an advantage.  My 12th best best match hasn't been dld in 18 months and my 15th in one year (35 and 43 V/D).  

It's really curious how some of these old and unsuccessul images get a good best match position, but I am not so convinced that V/D are the key to best match.

Regards,
Adelaide


« Reply #15 on: November 10, 2008, 18:15 »
0
My two best-sellers are one in the first page, the other on the last. Almost iddentical dl/views ratio, uploaded with a month of diference time ago...
What people forgets it's that with all probability best match always include a random percentage, to diffcult reverse engineering.
That said--- I think this best match isn't so bad at all.

« Reply #16 on: November 10, 2008, 18:16 »
0

Sorry, I don't get it...

How could your portfolio be sorted according to a Best Match; best matching what? There must be a reason why my photos shows in a given order when I look at my portfolio, but I think it has nothing to do with a search with keywords.

Am I wrong?

Claude

best match gives each file a value depending on whatever the current criteria are, relevancy to a search term is not one of them, so when you look at your own portfolio by best match  you see it ordered in ranking according to the current formula.

« Reply #17 on: November 10, 2008, 19:48 »
0

best match gives each file a value depending on whatever the current criteria are, relevancy to a search term is not one of them, so when you look at your own portfolio by best match  you see it ordered in ranking according to the current formula.


Yes, but don't you think that these ranking criteria must drop when the search by keyword comes in?

Claude

vonkara

« Reply #18 on: November 10, 2008, 20:09 »
0

best match gives each file a value depending on whatever the current criteria are, relevancy to a search term is not one of them, so when you look at your own portfolio by best match  you see it ordered in ranking according to the current formula.


Yes, but don't you think that these ranking criteria must drop when the search by keyword comes in?

Claude
I don't see how it could drop? If you enter a search, all the images with this keyword comes up. Then it's sorted by the most relevent using the best match criteria (dwl/month... views..). I don't see how to rank all these files with the same keyword with only keywords?

« Reply #19 on: November 10, 2008, 20:26 »
0
I think the best search criteria of all is one that does not give preferences to any particular group  because of their status within an organization such as being exclusive or because of their canister level. This has totally and absolutly nothing to do with what the buyer wants. I think the search parameters should be at its simplest form such as the one at Shutterstock. Search by newest, oldest,  random or most popular. Most popular is simply ranked by Number of Downloads divided by Number of days online,  regardless of who you are.  Why should it be more complicated then that? Denis
« Last Edit: November 10, 2008, 20:33 by cybernesco »

PaulieWalnuts

  • On the Wrong Side of the Business
« Reply #20 on: November 10, 2008, 20:42 »
0

best match gives each file a value depending on whatever the current criteria are, relevancy to a search term is not one of them, so when you look at your own portfolio by best match  you see it ordered in ranking according to the current formula.


Yes, but don't you think that these ranking criteria must drop when the search by keyword comes in?

Claude

Think of it as a general ranking without factoring in keywords. It's an overall gauge.

PaulieWalnuts

  • On the Wrong Side of the Business
« Reply #21 on: November 10, 2008, 20:58 »
0
The LAST PICTURE of my portfolio has 16 DLs and 118 views. QED. Of course now it will neither be seen nor downloaded. These f****** idiots at IS. Note to self: Do not upload highly specialized pictures that target a tiny market because if they fit too well they get KILLED. Morons.

Maybe you should extend your opprobrium to those who were gaming the system this time last year with their buying gangs shooting a selection of seasonal images right to the top of the best match - self perpetuating incredible dl/view ratios. No doubt this best match is over reactive to that scenario at this peak buying time but moronic behaviour is not the sole prerogative of istock.

I don't think Istock is intending to generate bad results. They most likely are facing several issues they're trying to resolve and while they're getting a positive result for one issue it may have a severe negative affect on another.

I was new to IS this time last year so I didn't notice the gaming. It would make sense that files which have a very high view/dl ratio get bounced down in ranking but here's the problem with that. It's punishing decent people who have good sellers. And, the gamers will most likely always figure out what moves IS is making and adjust accordingly. Are there that many gamers that they just can't run a best match dl/view analysis and single out contributors who have a suspiciously high ratio? Is it necessary to punish individual files that perform well?

« Reply #22 on: November 10, 2008, 21:30 »
0
I don't see how it could drop? If you enter a search, all the images with this keyword comes up. Then it's sorted by the most relevent using the best match criteria (dwl/month... views..). I don't see how to rank all these files with the same keyword with only keywords?

I get it! 

claude

lagereek

« Reply #23 on: November 11, 2008, 02:50 »
0
Agreeing.
I dont think the best match is there to purposly generate bad results and ofcourse if its going to hit some category bad well obviously it would be the non-exclusives. Fair enough.
The problems within the site seem to be many, which doesnt really inspire confidance either. It all started when Getty foeced their CV onto IS.
The Getty CV, just take a look at it! its for creative-buyers, browsing after RM,RF images. Their CV has got absoloutely nothing to do with Micro, where people are buying shots off the peg.
I cant for the life of me understand why the IS Admin went along with this.
Ive always had my own little 6 or 7K picture library, extremly specialized towards the industrial Power-generation, leasing out the shots for ADs and Annuals etc working like a private-client basis. I.E. regular clients. Ive done this little side thing for 20 years totally aside fron general Stock. When they discuss a brief with me they use the terminology pretty similar to that of Getty, you know, standing-shot, room for falling text, nobody, somebody etc. Why? because their end result will be a big elaborate printed matter!!
The typical IS customer is somebody who wants a little webb shot!
Thats why IS should have kept a lightyears distance away from the Getty CV.

Sorry guys, computer and webb knowledge simply isnt enough in this business.
Youve got to THINK like the market your aiming at.
« Last Edit: November 11, 2008, 02:55 by lagereek »

bittersweet

« Reply #24 on: November 11, 2008, 07:56 »
0
Agreeing.
I dont think the best match is there to purposly generate bad results and ofcourse if its going to hit some category bad well obviously it would be the non-exclusives. Fair enough.
The problems within the site seem to be many, which doesnt really inspire confidance either. It all started when Getty foeced their CV onto IS.
The Getty CV, just take a look at it! its for creative-buyers, browsing after RM,RF images. Their CV has got absoloutely nothing to do with Micro, where people are buying shots off the peg.
I cant for the life of me understand why the IS Admin went along with this.
Ive always had my own little 6 or 7K picture library, extremly specialized towards the industrial Power-generation, leasing out the shots for ADs and Annuals etc working like a private-client basis. I.E. regular clients. Ive done this little side thing for 20 years totally aside fron general Stock. When they discuss a brief with me they use the terminology pretty similar to that of Getty, you know, standing-shot, room for falling text, nobody, somebody etc. Why? because their end result will be a big elaborate printed matter!!
The typical IS customer is somebody who wants a little webb shot!
Thats why IS should have kept a lightyears distance away from the Getty CV.

Sorry guys, computer and webb knowledge simply isnt enough in this business.
Youve got to THINK like the market your aiming at.

I'm not sure I agree with this statement, since clearly there are a lot of images being sold that are not little web shots, and it's been demonstrated that the istock customer comes from all markets, including huge corporations.

What I believe is the real problem is that Getty's CV works perfectly on Getty because the keywords are being assigned by editors who are not emotionally attached to the images and lack any desperation whatsoever for whether the file ever sells. On the flipside, the keywords at istock are entered by contributors who, as demonstrated repeatedly by some of the laughable complaints posted on this very forum, are sometimes borderline delusional about what words a designer might use to find their image AND actually be within the realm of what they hoped to see.

Misguided misunderstanding and/or malevolent intent by contributors who care about nothing but their files showing up in as many searches as possible--regardless of whether they are actually relevant--are responsible for the CV failure at istock.

lagereek

« Reply #25 on: November 11, 2008, 08:59 »
+1
Misguided delusional contributors spamming is a differant issue. Since IS didnt have the guts to throw the book at them, early stage, well there you go. Some 25% of IS shouldnt be penalized because IS dont sort out the keywording properly BEFORE images get accepted.
Believe me Ive been with the Getty-RM since 1990 and their general customer is very, very differant from the IS punter. Sure IS also get some creative clients but bottom line is: shots off the peg, make no mistake about that and its absoloutely nothing wrong with that.


bittersweet

« Reply #26 on: November 11, 2008, 09:03 »
0
Since IS didnt have the guts to throw the book at them, early stage, well there you go. Some 25% of IS shouldnt be penalized because IS dont sort out the keywording properly BEFORE images get accepted.


I agree with this 100%.


 

Related Topics

  Subject / Started by Replies Last post
41 Replies
16408 Views
Last post September 14, 2011, 17:43
by madelaide
10 Replies
3167 Views
Last post December 13, 2015, 17:04
by weathernewsonline
25 Replies
4690 Views
Last post February 29, 2016, 23:39
by AlessandraRC
14 Replies
5818 Views
Last post November 07, 2016, 14:32
by CJH Photography
1 Replies
1997 Views
Last post January 08, 2017, 23:08
by AnTiImAgeX

Sponsors

Microstock Poll Results