pancakes

MicrostockGroup Sponsors


Author Topic: Is what you are currently doing working?  (Read 16196 times)

0 Members and 2 Guests are viewing this topic.

« Reply #50 on: October 10, 2017, 13:00 »
0
JustinB, how can you say "I am working toward a solution for our collective future" while at the same time I see images from sites like DepositPhotos are high up in your search?  Perhaps  lucky for us that Picturengine doesn't appear to be much use for buyers.  I think it ruins your credibility, you can't claim to be trying to help us while helping sites that make it unsustainable.


PicturEngine currently searches over 850,000,000 organic images.  We will cover the entire industry, giving image buyers a comprehensive selection of stock photos available for license.  Some images rank higher than others in our search results; this will change as more buyers train the system to what they like seeing in their individual search results.  I have discussed several times how we list images in our search results, Organic vs. Paid.

Do you stop using the searches from Google or Bing (or other search engines) because they show you a particular website or search result you dont like?  They probably provide that search result because of your past searching habits.  Click on more images that you DO like.

Sites like Alamy and Pond5 keep the industry sustainable for me.  There's others that pay 50% but they need more contributors and buyers to back them.  PicturEngine promotes sites that pay us very little and makes the industry more unsustainable.


I believe stock photo agencies should pay fair commissions to creators.  My agencies Picturesque and Corner House Stock Photo paid 50%, and now photographers receive 100% using the PicturEngine platform.

I do think there will be a better way to sell image licenses in the future.  Blockchain technology looks interesting.  I think we will have a very low fee option one day without having to pay middlemen.  Until then, sites that pay 50% will do for me.


Blockchain does look interesting, however by itself it is merely a ledger of transactions written to a permanent block.  Cryptocurrencies that use blockchain for transactions such as Bitcoin do have transaction fees.  This is how the distributed blockchain pays for itself (and pays the miners that make the operations possible).  These fees are based on many variables and often have long delays coupled with currency fluctuations. In my opinion too risky for now.
http://mashable.com/2017/08/28/bitcoin-transaction-fees/

The PicturEngine Platform currently uses PayPal (set fees we can predict) we can switch (or add) any widely used transaction method in the future.  It is important to me that PicturEngine does NOT touch your transaction/money; it passes straight from buyer to seller instantly when an image is licensed.


GraniteCove

« Reply #51 on: October 10, 2017, 17:41 »
+2
The concept makes me think of Fotosearch with a payramp for individuals. Obviously Im skeptical, partly because the OP has yet to answer any of my questions, but also because I m fully aware of the kind of resources it would take to establish a truly credible and competitive alternative to the big agencies. But his post did get me thinking about a strictly pay per play agency model and whether or not I would even consider contributing to such a beast.

My initial thought was absolutely not. Why would I? But then the wheels started turning and Im beginning to see that maybe some kind of pay to play model might not be such a bad idea after all. A more or less collective agency something along the lines of Stocksy (but without the entrance barriers) where a nominal PER IMAGE upload fee could potentially provide considerable seed and ongoing marketing capital for the good of all. The per image fee would also in my opinion force each of us to edit more carefully and upload only the best of our best images, eliminate needless similars, and (I know Im going to get flak for this) in short order winnow out the serious contributors from the chaf.  In essence creating a high quality, self curating library with real potential for commercial traction and buyer acceptance. Like Stocksy too, image exclusivity would be mandatory but not contributor exclusivity; leaving contributors free to continue uploading their lesser images elsewhere.

As a contributor the trade off for my up-front risk would need to more than offset the cost of contributing, so one of the things I would need to see would be a no subscriptions model to start coupled with a prudent  70/30 or even a 60/40 split in my favor.  Whenever I see 100% for you the contributor, I just know its going to fail. Again, like the Stocksy model  I personally dont believe pricing needs to be as low as the big agencies to be competitive but on the other hand not quite as high as theirs either. Somewhere mid-point would still be highly competitive considering the market a site like this would be targeting. I would also need rock solid assurances that my upload fee would go directly toward ongoing marketing of the site. A tall order I know, and I dont see this sort of thing coming any time soon necessarily, but one thing is for sure - something needs to change because the current model doesnt seem sustainable for much longer. Not for contributors like me anyway.

« Reply #52 on: October 10, 2017, 23:18 »
0
"Yes, your images are probably already in our organic search results if they are listed with an agency.  Using our paid Advertising or Platform replaces your images within our search results, and sends the buyer straight to your site or selling page where you make/keep more money."

I think the project is intriguing on a number of levels but after looking at your pricing plans I am confused by your business model. As mentioned in a previous reply, offering lower rates to agencies than to individuals does seem to contradict one of your expressed goals. The real confusion for me though is in your advertising only packages. For instance, let's say that I as an individual wanted to leverage your advertising to drive sales to a specific stock site that offered higher commissions or to increase my ranking there, how does your algorithm handle conflicts between two or more paying advertisers? In other words, if your platform gains popularity to the point where it disrupts agency sales what is stopping agencies from advertising alongside contributors. Since the majority of us contribute to multiple agencies how exactly would that work? Furthermore, how would/does it handle conflicts between paying contributors for similar searches?


First, see our public FAQs
http://picturengine.desk.com/customer/en/portal/articles/524043-am-i-competing-with-my-agencies-
http://picturengine.desk.com/customer/en/portal/articles/478867-does-picturengine-compare-prices-

When an image is ingested into PicturEngine, we start with a deduping process (removing or ranking the exact duplicate image.)  Our industry is full of duplication BOTH from agencies and photographers syndicating (sending the same image to multiple sources.) It is important that we do not allow agencies to compete on price, so we only want to show the image once in our search results. 
We use a scale from 0-10.  With "0" represented as licensing directly from the creator/copyright holder, either on our PicturEngine Platform or Advertising Only (on the photographers' sales and delivery platform.) When an image is marked as direct to the creator "0" we default and only show this image to the buyers. If an agency is paying for advertising, the lowest they can get is a "1" on this scale.  We use a formula to try and determine the "base" agency for a particular image the lower number on the scale gets shown to the buyer.  We had to come up with a system because we found the same image at several agencies and also direct from the creator.  I don't want to make a super long post, let me know if you need further explanation, (there are many calculations that go into deciding the 1-10 ranking of agencies on a per image basis.)

At the moment it looks as though your site only returns one page of results for any of the search terms I input, however what happens when that expands to ten, or a hundred, or a thousand? I know if I am paying the same amount for exposure as my competitors then I don't want page 20 or even page 2. I want number 1 page 1. So then what? Sort by popularity, relevance, new? I get that for free already. Maybe by rotation then? I suspect I get that for free already too, so the only thing I can see if the platform takes off is a pay for placement auction scheme similar to Google adwords. Is that correct?

Forgive my skepticism but I like to know what I am actually getting before I spend my money. By the way, are you aware your splash page explicitly states "PicturEngine features unbiased searches"? You should probably remove that.


The search results are listed as an infinite scroll. When you search for a particular term, you receive a set of results; we load about 100 images at a time.  As you (the buyer searching) act on the search results, click on images, add images to a lightbox, etc. our system is listening and learning what you like to see in your search results and can then suggest images in the next load of results as you scroll.  Think of our search as a big funnel starting wide at the top and narrowing as you act on the results and we learn what you like seeing.

The goal is a smart, learning image search where no one can pay to have higher results (no auction type mechanism, as that, won't benefit the creator or buyer) Instead we try and determine the best result for that person searching.  Giving a shuffled result of both organic and paid results giving the buyer the best search experience possible. 

On the flip side, we also want to give creators data as to how their images are performing, and what they can do to improve.  The good images (for that particular search/subject) should rise, and the not so good ones fall, let the buyers actions decide.

We are always interested in feedback that's why I started this thread.

« Reply #53 on: October 10, 2017, 23:40 »
0
The most successful companies focus on what customers want, not suppliers.
That is the plan.  Focus on a great image search.  We have received surveys and feedback from over 16,000 paying image buyers. 

We also listen to suggestions and ideas from photographers and agencies with what they want.  It is a balance.

I am doing this to solve the problems our industry is facing, adapt and adjust until we have a winning formula.  Thanks for your feedback and support.   

« Reply #54 on: October 11, 2017, 02:39 »
0
I'm still unimpressed by the results from the searches I did.  It took too long for the second page to load.  It did seem like there was only one page of results.  Then I had no idea how many pages there would be.  There were still less images than on single sites, like Alamy.  It was a bit quicker when I tried it just now, so maybe the speed varies through the day?

It would help if you put the total number of images found on the first page but then I suppose that makes buyers realise there isn't as much to look at as there are on some sites and your USP for them has gone?

GraniteCove

« Reply #55 on: October 11, 2017, 07:41 »
+2
Thank you for trying to answer my questions JustinB, , and trust me, I did read your FAQ's before I posted. The answers to my specific questions were not there (nor were they in your most recent post) so yes,  I do require further explanation.

My questions were not related to how the system handles non-conflicting queries, I understand how that works,  but rather how it reconciles searches from competing advertisers on your platform. Perhaps I was not clear enough so I will reframe the question. What happens when you have say 100 photographers with photographs of lets say a pizza. Let's also assume that they are all similarly keyworded which they would naturally be. Since all the photographers are paying  the same advertising rate to you how on earth does your system even begin to prioritize those images in the initial sort?

Another question was related to how your system would deal with a photographer who was not interested in selling directly necessarily, but rather leveraging your advertising platform to drive sales to a specific stock agency (something I think could be very useful). I understand how that would work in your organic vs paid scheme, but I don't quite get how that would work if the agency I wanted to direct sales to also advertised with you. Or to complicate things even further, multiple agencies with the same image. Who exactly gets priority and how does your system decide who gets dumped from the search? Either way, someone is not going to be happy.

Please go back and reread my first post.

« Reply #56 on: October 11, 2017, 23:46 »
0
It would help if you put the total number of images found on the first page but then I suppose that makes buyers realise there isn't as much to look at as there are on some sites and your USP for them has gone?

Providing a count of total images from a search that learns, adapts, grows and shrinks as you act upon it, was not practical.  The image results count (or totals) continuously change as you act, click, etc., the search results are not supposed to end until you find what it is you are looking for. 

As more users search, click, act and train our engine, smarter (better) it gets at inferring the meaning in words entered to search.  We are not there yet, but that's what we are striving to create. 

« Reply #57 on: October 12, 2017, 14:09 »
0
My questions were not related to how the system handles non-conflicting queries, I understand how that works,  but rather how it reconciles searches from competing advertisers on your platform. Perhaps I was not clear enough so I will reframe the question. What happens when you have say 100 photographers with photographs of lets say a pizza. Let's also assume that they are all similarly keyworded which they would naturally be. Since all the photographers are paying  the same advertising rate to you how on earth does your system even begin to prioritize those images in the initial sort?

I understand now; you want to know the mechanics of ranking very similar images for a particular keyword, subject, all paying to be in the search results. For this hypothetical let us assume 100 Platform or Advertising Only photographers submit pictures of pizza, all of the images are similar but still have distinguishing factors or different lighting and all of the pizzas.  For ranking these, we look at many of facets to determine an images' rank and order.  Besides the obvious, what buyers like seeing and have clicked on in the past; color, focus, orientation, copy space, camera view, license type, also other keywords, descriptions, and categories, attached to that image.  We also take into account that photographers particular specialty and what else is in their collection submitted.  All of these facets are evaluated, and each is assigned a score.  This is a fraction of the list we use to evaluate an image.

It becomes a very hard problem to solve indeed when it gets to the point of 100 photographers with 100 cheese pizzas images, all very similar in all respects, all shot from above, with no other distinguishing factors or facets... after all else is exhausted for deciding an images' ranking, in this extream example (for now) we randomize.  That will likely change as we discover more facets to compare. 

Another question was related to how your system would deal with a photographer who was not interested in selling directly necessarily, but rather leveraging your advertising platform to drive sales to a specific stock agency (something I think could be very useful). I understand how that would work in your organic vs paid scheme, but I don't quite get how that would work if the agency I wanted to direct sales to also advertised with you. Or to complicate things even further, multiple agencies with the same image. Who exactly gets priority and how does your system decide who gets dumped from the search? Either way, someone is not going to be happy.

Please go back and reread my first post.

If the image belongs to a photographer licensing directly, they are automatically the default 0 position.  Ranking paid agencies on our scale of 1-10 (as described in my earlier post) is no easy task. We started by using first (oldest) upload date; this was to benefit early adopters to our Advertising Only for agencies. 

We developed this strategy of first come, first served, early on for simplicity in billing. Example: Agency A pays us to list 300,000 images.  Agency B signs up next and also has 300,000 images however 200,000 of this collection are the SAME exact images as Agency A, plus 100,000 unique images.  Agency B only gets billed for 100,000 images because they are unique.  So hopefully not too many feelings are hurt.  This benefits early adopters.  The way I see it, if an agency is paying for advertising in our search results, they are trying to get those images licensed to benefit themselves and those photographers they represent.  So they are working for the photographers' best interest and should be rewarded.
 
This can change in the future if photographers tell us an agency is not treating them fairly, not paying commissions on time, etc.  Photographer feedback is important to us. 

GraniteCove

« Reply #58 on: October 12, 2017, 18:33 »
0
I appreciate the answers JustinB. It clarifies things for me.

« Reply #59 on: October 16, 2017, 17:46 »
0
I think this is interesting and relevant to our conversation about the search engine business model we are having on this thread.  Over the years Google has added more paid advertising space and less organic to its search results pages. 

https://moz.com/blog/google-organic-clicks-shifting-to-paid

This practice is all too common in the world of search engines, and it teaches us that paid content will in the future be the only way to be found.  Organic site content that comes up will only be the most significant and specialized in that particular subject. 

This is precisely why we do not want to allow a bidding ranked system on PicturEngine for our paid results.  I want to let the image buyers choose what they want to see in their specific search results.  Why I decided to spend the resources on our suggestion engine and learning image search.  We want the best image for that buyer to come up in their results, not the users' that can pay the most.

The world is changing around us.  We need to be agile and adapt to that change.  Thank you all for your suggestions on how to make the industry more sustainable.  Please keep them coming.   


 

Related Topics

  Subject / Started by Replies Last post
39 Replies
17552 Views
Last post March 30, 2009, 00:08
by sgcallaway1994
25 Replies
13012 Views
Last post April 01, 2009, 06:15
by CofkoCof
HEY... It's Working

Started by WarrenPrice Cutcaster

19 Replies
10226 Views
Last post February 23, 2011, 17:15
by stockastic
1 Replies
4739 Views
Last post January 29, 2013, 13:11
by 08stock08
1 Replies
5964 Views
Last post February 09, 2014, 06:38
by Beppe Grillo

Sponsors

Mega Bundle of 5,900+ Professional Lightroom Presets

Microstock Poll Results

Sponsors