Author Topic: That time of the month - Percentages for July (Read 31601 times)

lisafx · « **Reply #75 on:** August 07, 2010, 10:56 »

Quote from: FD-regular on August 06, 2010, 18:34

Last year I did correlation statistics between downloads and uploads over 3 years. At least for me, the correlation was significant. (the method was based on time series covariances).

What did the correlation statistics show?

FD · « **Reply #76 on:** August 07, 2010, 12:40 »

Quote from: lisafx on August 07, 2010, 10:56

What did the correlation statistics show?

That there was a clear covariance between uploads and subsequent downloads. It took me a lot of time since I had to go back over 4 years to fill in all the uploaded batches. The variation was good enough since I had some periods of no uploads at all, and periods of high uploads.
Caveat: this is about the past. If they are in beta for a new search engine, it might all change.

With the risk of being a bore, I had to use that method in one my PhD sections in order to demonstrate a lagged effect of food intake on gut hormones release. Basically, you do a normal Spearman Rank r between the two datasets while shifting one dataset day by day. I think the optimum was 5 days later but I didn't check for weekend artifacts. To be really valid, this should be done over larger ports than mine and from several people.

Kone · « **Reply #77 on:** August 07, 2010, 13:33 »

Quote from: FD-regular on August 07, 2010, 12:40

Quote from: lisafx on August 07, 2010, 10:56
What did the correlation statistics show?
That there was a clear covariance between uploads and subsequent downloads. It took me a lot of time since I had to go back over 4 years to fill in all the uploaded batches. The variation was good enough since I had some periods of no uploads at all, and periods of high uploads.

Thanks for this FD, I doubt that you will have a steady income if you do not upload regulary.

lisafx · « **Reply #78 on:** August 07, 2010, 14:19 »

Thanks very much for posting this FD! Looks like a lot of work went into it.

If I am reading your chart correctly, it looks like rate of uploading has a more direct impact on earnings than overall portfolio size does.

WarrenPrice · « **Reply #79 on:** August 07, 2010, 18:09 »

Quote from: lisafx on August 07, 2010, 14:19

Thanks very much for posting this FD! Looks like a lot of work went into it.

If I am reading your chart correctly, it looks like rate of uploading has a more direct impact on earnings than overall portfolio size does.

Lisa (or FD)
By "rate of uploading" are you talking about frequency or quantity? I'm uploading regularly, just not very much. $:-\$

luissantos84 · « **Reply #80 on:** August 07, 2010, 20:14 »

This chart is regarding all agencies? It is a portfolio average?

FORGET! SS ok... I have never stop the uploading there but once I have a week without uploading I feel it for sure..

FD · « **Reply #81 on:** August 08, 2010, 02:45 »

Quote from: lisafx on August 07, 2010, 14:19

If I am reading your chart correctly, it looks like rate of uploading has a more direct impact on earnings than overall portfolio size does.

Yes, well, both seem to be important. There are a couple of caveats though. Earnings in general seem to follow portfolio size up till late spring 2008, then hit a ceiling. I saw exactly the same pattern on BigStock. Some people here reported this glass ceiling too, but later.
It makes me suspicious that this is (also?) a side effect of my portfolio composition since in May 2008 I settled down; my travel/nature shoots stopped and I started models. With model concepts there is the problem that (a) you have a lot of competition from much better photographers and (b) your best concepts are copied like hell. I'm also handicapped by the age group and ethnicity of my models, mainly because they are my students and Asian. [My top sales on SS are still niche landmarks (difficult to copy) like "Waterloo battle field" where I dominate the top 20, or "Chinese sweatshop" where I'm alone.]

A second caveat is that the correlation has been done over the full 4 year period. People here reported around end 2009 that new uploads didn't seem to work as good any more on SS like before. You can see that at the end of the graph. But the correlation is of course over the full 4 year period and just one statistic (number) so changes in time are not accounted for. I'm hesitating to do the analysis over the past year since I had steady uploads, and obviously you can't do a proper correlation then, since there will be no variance in one of your datasets. [example: suppose you do a study about success in business and body height: you will have zero correlation if all your subjects are 5ft9]

A note about my method (for Luis): samples (rows) were per month and filled in an Excel spreadsheet. Columns were: earnings (light blue graph) and uploads per month (count # of images over all approved batches that month) (dark purple graph). The port size (light yellow graph) was just a cumulative sum of the uploads.

Conclusion: to have real hard evidence about the "beast feeding" hypothesis you need to have several portfolios (to get rid of the port composition bias) and larger ones. Also, the past is no guarantee for the future since SS can change the algorithm as they like.

luissantos84 · « **Reply #82 on:** August 08, 2010, 08:28 »

thanks

I would do the same but my stock experience is really small and I have never stopped the uploading but I repeat that I feel the same about SS, but actually is the only one that has nice earnings

MicrostockExp · « **Reply #83 on:** August 11, 2010, 04:10 »

A bit late, but here are my July stats for photo/video/audio:
http://microstockexperiment.blogspot.com/
Cheers
L

BaldricksTrousers · « **Reply #84 on:** August 11, 2010, 05:16 »

Quote from: microstockphoto.co.uk on August 06, 2010, 04:50

Quote from: gostwyck on August 05, 2010, 11:43
Maybe this is when the going gets tough and we see who is tough enough (or foolhardy enough) to keep going.

I've got a feeling that we may count many of us - which are actively posting in this forum - in that number. Who else?
Not necessarily bad in the long term.

I've thought for a long time that falling returns per file coupled with rising quality demands would ultimately freeze out new contributors and most of the long tail, leaving a few hundred or a few thousand of us sharing the cake with portfolios that either become very large or are have very high commercial value, or both. The sooner that happens, the better.

In the summer of 2004 I was making between 80c and $1 per file per month on iStock, with a tiny portfolio. Today I make just 15c per file per month there (about 50c per file from all sites combined - of course, it would be about 45c on iS if I went exclusive ... but for a newbie the non-exclusive rate would apply), so the return has halved while the effort has multiplied several-fold.

At the same time, I have had to learn a hell of a lot about photography, just to get where I am, and I've also invested more than $10,000 in cameras, lenses, lighting gear, computers etc. . It has all turned out very satisfactorily for me but I think it would be insane to contemplate doing it seriously, starting out at 5 uploads a week on iS with two or three likely to get rejected so you build your port at a rate of 10 files a month and maybe hit bronze after two or three years. SS is better, of course, but that's the only site offering a decent chance of a return without restrictive upload rules.

Even so, it seems likely that the only people who stand a chance of breaking through these days are those who really understand the stock business and are already highly proficient photographers or illustrators, maybe with a back catalogue of useful images. And it's still going to be tough to make an impression.

Am I wrong about this? Would any of us like to be starting from scratch now?

FD · « **Reply #85 on:** August 11, 2010, 07:35 »

Quote from: BaldricksTrousers on August 11, 2010, 05:16

Am I wrong about this? Would any of us like to be starting from scratch now?

I think I even wouldn't get in any more at SS and IS. In 2005 it was much easier, and the technical requirements were less. In 5 years, suffering rejections and learning what sells and doesn't sell, we oldies have grown with the sites. If you want to enter now, you need to have that skill acquired outside stock or you can't pass the high threshold. There will always be fresh photographers with loads of talents, but the avalanche of 1-2 years ago is over, I guess.

MicrostockGroup Sponsors

Author Topic: That time of the month - Percentages for July (Read 31601 times)

lisafx

FD

Kone

lisafx

WarrenPrice

luissantos84

FD

luissantos84

MicrostockExp

BaldricksTrousers

FD

Related Topics

Sponsors

Microstock Poll Results

Sponsors