Quote from: Lee Torrens on March 12, 2010, 09:41

A note on the benchmarking data:  We didn't want anyone to be able to maliciously influence it. So we came up with the concept of 'trusted accounts'.  I contacted some friends and pulled in a lot of favors to get people I knew and trusted to participate and input their data. I then marked their accounts as 'trusted', which puts their data into the pool of benchmarking data.  There's a healthy range of contributors, from top to bottom in both portfolio size and selling success.  To all those who did that for us, thankyou very much!  

Sorry, this was a bit confusing then.  So, the only statistics people will be able to compare themselves to is this relatively small group of "trusted" people.  And anyone's inputted data will not be used for any purpose aside from their own needs.  Got it.

Ooops I did not get it in first place too.

Alternatively you can use baselining (http://en.wikipedia.org/wiki/Baselining ) but it would require some time to run it before you gather enough data to have baseline.

Quote from: melastmohican on March 12, 2010, 12:49

From usability perspective it is much harder to use than spreadsheet where you can have a table in which you can input all data at once. It's not very fast when you have to click around to switch pages to enter one number per agency.

Agreed. We've got a lot of work to do to make it easier, and we may never arrive at a point where it's superior to a spreadsheet in terms of usability.

Quote from: melastmohican on March 12, 2010, 12:49

On the bright side, when you collect  all the data from us (it means we do all hard work) you can do a lot nice statistical things. No mention that you can sell it to third parties. So overall it's great idea derived from LookStat fiasco when they could not collect data automatically cause agencies do not want to share it across industry. I wonder why iStock have API to share statistics and others do not want to that.

Please don't do "all the hard work" if you don't see the value in it.  We're only using the data from 'trusted accounts', so we don't benefit from more people using the service, aside from traffic on the site to help sell advertising. 

Our initial business model included selling the combined and anonymous data to third parties, but we opted to make that data public and free and support the costs with advertising instead. We think that helps the service have a better position in the industry as everyone stands to benefit. We'll see how things go with the advertising and referral links model, but for now we have no plans to sell the data.

The idea wasn't so much "derived from LookStat fiasco".  I actually shared this idea with Rahul over dinner in New York in 2007, well before LookStat was even the brand of Rahul's business. But yes, we always intended to avoid scraping data from agencies. That's one of my fundamental ideas. If the agency doesn't give me the data willingly via and API or XML feed, I won't take it.  That being said, I'm not opposed to leveraging the data from others who do so.

Finally, the iStock API isn't so much about sharing statistics. Understandably, they only permit use of their API when it's in their interests.  If you're referring to iStockcharts.de, that site scrapes data from the website pages and has never had official support from iStock management. iStock do share more statistics that most agencies though, which is generous, but also in their best interests.  If it wasn't they would have taken it off rather than 'blur' them with the >100, >1000 strategy.

Lots of interesting topics for such a small comment.  Thanks for the input. 

In some other thread somebody mentioned import/export which would be helping too.

I think value would come from the large volume of data. So if it is going to be limited to 'trusted accounts' (what ever it means: people from this forum, large contributors, etc) it's kind of pointless.  Regardless of the method of financing this venture (advertising or selling data) success would depend of volume of data to mine for useful statistics.In regards of agencies contribution to statistics I would say it is against their best interest. Main reason for that is that reliable data would prove that it's not profitable for most contributors and such data may stop flow of newbies trying to get rich quickly on microstock. Nobody actually can estimate how much they get from accounts that never reach payout limits. It;s kind of Ponzi scheme, it works only if there is constant flow of new members.

Recently SS announced that they are going to provide RSS feeds of individual portfolios. I tried and it works great. You got nice image previiews in a feed.

I tried to check if other agencies got similar feed and it does not seem to be very common
1. Fotolia seems to have it but there are no previews, just title and link to image. For some reason it cannot be imported to FriendFeed but it works with Google Reader.
2. IStock - useless. Feed points to main site not your portfolio
3. 123RF - same story as above. I cannot figure out how to have feed of my portfolio.

I have been uploading same stuff to all agencies. 123rf used to be very liberal in approvals so I have the biggest portfolio there which translates into larger sales than DT or FT. Recently I stopped uploading completely cause rejections went sky high there but still volume of images is twice as bigger that on any other agency. I am expecting that it will be harder to get images accepted across the board so people who were able to build their portfolio long time ago will still have advantage over newbies like me (less than 2 years in microstock). On the other side is IS. I guess I am constrained by upload limits but if returns are usually highest across agencies. Improvements are extremely slow but now they slowly advancing to my top 5. In next two years they might be my #2 with this growth rate or if they eat 3 others that are above them :-)

From usability perspective it is much harder to use than spreadsheet where you can have a table in which you can input all data at once. It's not very fast when you have to click around to switch pages to enter one number per agency.

On the bright side, when you collect  all the data from us (it means we do all hard work) you can do a lot nice statistical things. No mention that you can sell it to third parties. So overall it's great idea derived from LookStat fiasco when they could not collect data automatically cause agencies do not want to share it across industry. I wonder why iStock have API to share statistics and others do not want to that.

Great article. So far best I have seen on this subject. Not trying to just lure more people and earn money out of newbies :-) I guess realistic for most of the parts.

100,500 or 1000 it's still very low number in comparison to number of contributors worldwide.

anyone else plant trees for a living in University?

When I was in school, during communism, children were taken to site next to copper works to plant trees. Of course there were no filters installed so every year we came and have to start over :-) When it was raining acidic rain was burning thru girls stockings so I guess no living organism could survive there :-(

"The results from the third party  sales for fourth quarter are now in.
Your images made a total of �22,23.
This amount has been credited your account at YAY as one sale on a random image today."

Weird, do they actually know which photos are sold???

It was announced couple days ago. Firmware will be available in USA on March 20th.


http://melastmohican.net/wordpress/2010/03/02/eos-5d-mark-ii-firmware-upgrade-2-0-3/

I wrote bookmarklet that checks all checkboxes on a page and then opens new window for deleting checked images. Works on DT but I guess you can adapt it to work on FT.

http://melastmohican.net/wordpress/2008/12/11/got-tired-asking-for-simple-things-on-microstock-websites/

Great change. Still it does not show that for most of people top 4 is 10x more than next tier. It seems like earnings from SS are just 3x better than PantherMedia :-) Maybe you need more points at the bottom of scale which automatically would give higher weight to higher parts.

Now it makes sense why most of the footage sites required Photo-JPEG or Motion-JPEG codecs and reject anything which is MPEG4. These codecs are inferior but probably free.