Quote from: Mantis on August 29, 2014, 15:33

Hi Scott,

Does "upload timing" have more to do with seasonality? If not can you elaborate?

It could be due to seasonality, or a variety of other factors.   For example, "Christmas" search traffic begins in September or October, so you need to decide whether you want to capitalize on print publications or other paper-based products and retailers who might be putting layouts together months in advance, or digital buyers who are looking for images closer to the event itself.  Like any business in an open marketplace, you also need to think about how you're differentiating yourself from other sellers in the market.  If every other contributor is uploading Christmas ornaments in December, what are you doing differently in terms of timing, content or keywords?

Similarly, weight loss topics tend to peak in January as people make their New Years resolutions.     

If you're porting a full portfolio from another site, for example, and then put 3,000 images on all subjects up on the site all at once on a Friday afternoon in late July, that's not as thoughtful as spacing your uploads apart and trying to align peak customer traffic with differentiation and demand for your content. 

With 1 million customers, we often encourage contributors to treat the marketplace as a great place for doing their own testing.  Use Custom Sets and try uploading different topics at different times; pay attention to seasonality; watch Google Trends and try different topics out; space your uploads apart, etc.   Again, the contributors who do that tend to outperform those that don't. 

This list is very basic, but if you think about how some of those topics align with seasonality, that's the gist.

Again - sorry - I know some of this is basic, but it definitely helps.

Best,

Scott

My autumn photos started selling beginning of August.

This week my xmas and snow stuff started going.

It's the usual 2-3 month lead time for timed events.

Yeah, the trick there is that as you get closer to the event, search traffic can rise, but so can competition for a particular topic.  I think the key thing is to figure out how to best time your uploads, but also do something different / better, with robust and high quality keywords.  Customers can pick out things that are exceptional, different or unique, even within highly covered topics, and the quality keywording ensures that you've covered the bases.   It's amazing - you can see it in the download numbers - one image of [insert highly covered topic here] can get thousands of downloads while others of the exact same topic will get dozens.   It's because one contributor did more with styling, keywording, image quality, timing, etc..., and customers respond to that image over the others. 

The other important thing to note is that with 1 million customers - we're getting millions upon millions of searches.  There's still a ton of demand.  Some topics (Christmas, arrow icons, etc.) can be highly covered, but there are lots of others that aren't.


-S.

Holding back works too.  Get a lot of photos on a topic (example autumn).  Submit some to go live the beginning on August, keep the rest until the following year then do it again.

Far better that way until waiting until you get a chance to shoot autumn leaves then submitting them all at that time.

Oh, that sounds complicated. Will give it a shot and see if "timing' the shots right helps

BTW how possible is it that some of my rejects were due to the quality of my post processing rather than the quality of the image itself? Let me find some and post them to this forum

Quote from: scottbraut on August 29, 2014, 17:38
....It's amazing - you can see it in the download numbers - one image of [insert highly covered topic here] can get thousands of downloads while others of the exact same topic will get dozens.   It's because one contributor did more with styling, keywording, image quality, timing, etc..., and customers respond to that image over the others.......

I have heard this a couple of times, but honestly I think that a lot comes down to luck.

We see that a certain image racks up loads of downloads then try to analyze "why that image and not the other several thousand similar ones?" Using hindsight like that it's easy to come up with reasons that may or may not be true, or only partially be true.

Of course it has to be a quality usable image, but with the size of the library now there are many thousands of other images down in the searches of a similar quality and usability that aren't getting the downloads because that one happened to rack up several downloads initially. If you've studied randomness you know that strings of occurrences do happen more frequently then we would normally assume given their statistical improbability, the statistically probable only plays out as expected over the longer term. These images are promoted in the search after each of these random(ish) downloads, increasing their chances of future downloads.

If you are one of the contributors producing top end images getting one of these is increasingly like winning the lottery, rather than anything you can make happen, past creating content in this top tier. The idea that if we only study the top downloads enough we can come up with a sure fire formula for creating a winning image is just not true.