What Sells?

A fundamental flaw with stock photography as a business is that producers have little or no access to the data necessary to understand demand. In most any other business the producer has some idea of what is selling and how much demand there is for the particular product.

In order to plan future production shoots, stock photographers need to know more about what is selling. Not just in broad terms like “babies,” “business meetings,” “crowds,” “family recreation” or “animals,” but in specifics of what types of images in each category and ideally, the degree of demand for each.

Understanding the degree of competition in each category is also important, but creators can get some idea of that by simply doing searches on major agency websites for the subjects they are considering shooting. Through such searches they can determine the number of competing images in the category and how various creators treated the subject matter.



While such searches can be of some value, they can also be very misleading. We know that a huge percentage of the images in stock agency collections never sell. Some of the major agencies license rights to less than 1% of the images in their collections, annually. No one has ever wanted to buy the other 99%.

What about the one image caused a customer, or multi-customers, to choose it rather than all the others available. While the answer may not be simple the only way creators have a chance of understanding is to see those that have sold and compare those with the others that haven’t.



There is not much point in producing more like those no one wants.

When it comes to subscriptions a much broader cross section of images tend to get downloaded because subscription customers often use some of what they download for planning or reference proposes rather than in a finished product. Nevertheless, even in these situations some top producers find that no more than 10% to 20% of their production actually gets used. And that percent is declining.

If creators had a better understanding of specifically what might be in demand they could cut their costs, save themselves a lot of time and stop producing images no ones wants.

Why Don’t Distributor’s Supply This Information?

The distributors who have direct contact with the customers know what the customers are purchasing and have the data as to how frequently. However, for the most part they won’t share this information with their suppliers?

The argument they give is that they don’t want their competitors to know what they sell because the competitor might then encourage their producers to create something similar. That could cut into their share of the market. (This argument falls a little flat when we consider that most images are non-exclusive and a host of distributors all have the same images.)

In addition, some producers who have managed to stumble upon a best selling image or concept don’t want other producers (their competitors) to produce images similar to their winners. That could cut into future sales of their images.

So everyone holds their information very tightly and stumbles along.

The distributors feel that as long as more and more creators try to sell their images someone will stumble upon something that a customer wants to buy.

High volume production teams do have an advantage because they can learn from their own sales what situations work and which ones don’t. But, how many producers can justify producing 20,000, 50,000 or 100,000 images just to be able to learn what will sell and what won’t. And in the meantime 80% of the images they produce don’t sell. On top of this, there is no help for the little guys who make up more than 99% of the image producers.

Maybe It Is Time For A New Model?

In the early iStock days it was possible to do a “most popular” search on any set of keywords and see which images had the most DLs, how many DLs each had and the total images responding to that keyword. You could also check the number of DLs on the numbers 1 and 100 images and get a sense of how rapidly interest in the general subject category declined. By using different keyword sets, it was possible to get an idea of the demand for different subject categories. Many photographers got useful guidance on what to shoot from such searches.

Now, all that is gone. Customers can still do a “most popular” search, but it is unclear if the first image to come up is the most popular in all time, the last year, the last month or today. The algorithm is hidden. All that most producers can do is guess.

Many contributors may be happy that now no one can discover which of their images are good sellers. But, is that really to their, or their agency/distributors, long term advantage? Every production becomes a shot in the dark.

Agencies may need to consider whether it isn’t time to make more data publicly available. A database of just the images that have been licensed indexed on the number of times licensed would be very helpful. A keyword search of this database should return the images in the order of number of times downloaded and a number that shows where that image falls in the sequence of all images in the collection should be next to the image.

Thus, a search of a less popular subject might show the 1st image with number 408, 2nd with 563 and 3rd with 876. This would tell the searcher that there are a number of other image types that sell much better than this particular subject matter. Nevertheless, it would give the searcher a general idea of how popular this subject matter is relative to all other subjects and which images in the category are most frequently used.

There could also be search options based in the number of downloads in all time, in the last year and in the last month. This could help the searcher determine if there are any trends in the type of images within the category that are being used.

Such searches could also be very beneficial to customers hoping to identify trends of what is being purchased and used within a category.