There is a lot of talk about how Big Data will save the industry, but are the major stock image distributors using the data they collect effectively?
Given the huge number of
images currently in major databases, it would seem that a very high percentage of them are never viewed by anyone. If the distributors are collecting data properly, they should know which images are reviewed by customers and which aren’t. If there are lots of images that are never reviewed by customers, is there any way to generate revenue from those images.
Let me outline the problem in some detail and a possible solution.
Problem
When new images are added to a collection they are allowed to remain near the top of the search return order (SRO) for a little while. But the algorithm also needs to show proven good sellers and images that have been saved into light boxes early in the SRO. Consequently, there is a mixture of new images and proven sellers delivered for customer review.
We also know that customers almost never review all the images delivered in any particular search. They only look at a few pages and then move on with new search terms, or to a new site. It is believed in most cases customers will not look at more than 500 images delivered in any search before moving on. I discussed this to some degree in a
previous article.
With most searches on the major sites – Shutterstock, Alamy, AdobeStock, iStock, GettyImages – there are thousands, if not tens-of-thousands of images returned. One would think that the distributors would know which images are on which pages, and which pages have been opened. They may not be able to tell whether a customer actually looked at all the images on page 5, but they should know that the customer didn’t see any of the images on pages 6 or higher. All the images on page 5 and lower would be marked as having been reviewed on that date. All that is needed is the last date reviewed. One would hope that the distributors actually collect such data.
With such data, they should be able to go in and determine which specific images have not been reviewed by anyone in the previous year or two.
If an image hasn’t been reviewed by anyone in two years, then it would seem pretty unlikely that anyone searching this particular collection will ever see that image, let alone purchase it.
Then the question becomes is there any way to surface such images?
Solution
I would create separate databases – Shutterstock 2, AdobeStock 2, etc. Then I would remove all the images not view in the last 12 to 24 months from the primary database and place them in the new database. Customers who continued to use the primary database would not miss these images because they never look at them anyway. Having fewer images to deal with might even improve the efficiency of the search.
The secondary database would be available for customers looking for cheaper images. I might make the price for images in this database 30% cheaper than images in the primary database. I might also slightly raise the price of images in the primary database because now, in theory, this is a premium collection.
Whenever an image is licensed from the secondary collection, I would move it to the primary collection and treat it as a new upload as of that date. Thus, the image would have a chance to compete again at a higher rank in the SRO, and possibly generate more sales at the higher price.
If images are being licensed via subscription from the secondary collection, I might require that an image be licensed three or four times as part of a subscription before being moved back to the primary collection. Any image licensed once at a single-image price would be moved immediately to the primary collection.
Advantages
1 – Have collections at multiple price points, rather than a single set of prices.
2 – Surfacing some older images that are really of very good quality and exactly what some customers need.
3 – Able to adjust prices on each collection separately.
4 – Able to raise prices on some product without pricing other customers out of the market.
5 – The 30% separation may not be ideal, but that could be easily adjusted based on actual experience.
6 – Able to adjust dead time (1 year, 2 years, 3 years) in the primary collection before moving an image to the secondary collection.
7– System for generating revenue from dead images.
8 – More complicated, but certain popular categories of subject matter like New York street scenes, or soccer might want to get moved to the secondary collection sooner if the images are no longer being seen in customer searches. Subjects, like some scientific images that are only requested rarely may want to stay longer in the primary collection at the primary price level.