The big distributors are missing a huge opportunity to capture an even larger share of the market than they already control. Before detailing the opportunity, I need to outline the problem.
The Problem
Customers frequently want to license additional rights to an image they have used before, or one that they have seen used somewhere else. Often they don’t know the original source of the image. Even when a credit line is provided, and they know which collection to search, finding the specific image can be difficult given how large the databases have become. (Alamy has just hit 26 million.)
Good keywording helps, but often the images customers want to use have the same keywords as thousands of other images in the database. With keywords there has to be something very distinctive about an image that separates it from all the other images in its general category. Creators must identify those distinctive characteristic with keywords. If the creator forgot to add the proper word to the keyword list the image may still be lost in the pile. In most databases newer images that have never been licensed are shown first. Images published two or three years earlier are usually pushed to the bottom of the pile.
This is not to say that all customers are looking for images that have been used previously. Some want new, totally unique images that no one else has ever used. But, let me offer two bits of evidence that suggest that a significant number of customers are more than happy to use the same pictures others have used.
First, I want to take you back to the days of the print catalogs. In the 1990s a big agency might have 10,000 of 15,000 images in its print catalogs and hundreds of thousand of images in its general collection. In addition there were new images that didn’t make the catalog deadline coming in all the time. Despite the large choice available to the customer, the images that got used most frequently were those in the catalogs. And they got used over and over again.
Moving forward to the last few years we need to look at microstock. On most microstock sites it is possible to organize search returns based on the number of downloads. It is believed that a huge percentage of the customers use this method of search rather than the distributor’s default algorithm “best match.” The customer can also see how many times the image has been used by someone else. If a customer sees that the images has been used 20,000 times (as is the case with some images on iStock) does that cause her to turn away and look for something else? More often than not she will use that image rather than something that has never been used. In fact, most creators find that once an image has been downloaded 1,000 times it tends to get used much more frequently than other images.
So the question becomes, “How can the distributors make it easier for customers to find images that have been used previously?”
The Solution
If the big distributors – Getty, Corbis, Alamy -- were to offer customers a visual search option similar to the algorithm currently being used by
Google in their image search it would make it much easier for customers to quickly find images they want.
This visual search would differ from what Google is doing in that it would only search the distributor’s collection, not the entire web. It would also show all the images in the distributor’s collection that fit the search parameters, not just a small limited selection as is the case with Google.
Visual search is not foolproof. I’m not proposing that it be offered instead of the current keyword search. But if it were an option, it would provide a more efficient search experience for many customers and save then time.
If the image wasn’t there then the customer would know immediately that she needed to look elsewhere, or for something different. The customer wouldn’t have to waste a lot of time plowing through countless images that are of absolutely of no interest hoping to stumble upon the perfect image.
I would also suggest that within the visual search customers be given the option of 100% match or 95% or 90%. If the customer is only interested in an exact match to the image uploaded then she would use 100%. But, there are cases where a slightly different image with many of the same characteristics would also be of use. In those cases she might want to search for images that had most of the same characteristics with a few variations.
Use of such an algorithm should generate more revenue for big distributors. It would allow them to take additional market share from the small agency/distributors because customers wouldn’t need to go there anymore. Most of the medium size distributors and small agencies have some of their best, most in demand images represented on one of the big three, but these images tend to get pushed to the bottom of the collection either because of the size of the collection, or because in the case of Getty and Corbis they want to push certain imagery where they get a better royalty share.
One of the reasons customers go to the smaller, specialized collections is that it is easier to find what they need without having to wade through lots of images that don’t fit their requirements. If customers were able to as easily find the right image in a large collection, and know that they could get better discounts by going to the large distributor, a big part of the reason for searching small collections would be eliminated. Most customers are well aware that they get better discounts by going to large distributors than to the smaller independent companies.
To some extent this same strategy could benefit microstock companies and keep customers from running off to one of their competitors, but customers would not benefit from discounts in the way they would from traditional RF and RM distributors.
It is hard to understand why companies like Getty have not implemented this technology. They own PicScout. PicScout could probably implement this option for Getty in about five minutes. One of the disadvantages Getty might perceive is that it would be harder to force customers to use the images Getty puts upfront in its search return order. At some point customers will get tired of searching for the perfect image and choose the best option they have already seen rather than searching further. That’s what Getty and Corbis are counting on.
Another risk if the customer knows immediately that the image she really wants isn’t on the Getty site, she might immediately jump to Corbis or Alamy to see if they happen to represent the image. In such a case the customer wouldn’t continue to search Getty’s collection and settle for 2nd best. That would not be in Getty’s best interest, and of course “what’s best for the customer” is not first priority.
Masterfile offers an “Upload&Find” option. But the indications are that it is not widely used. However, there are some big differences between Masterfile’s collection and the collections of Getty, Corbis and Alamy. Masterfile doesn’t have anywhere near as big a collection so it may not be that difficult for customers to find what they want using keyword search. In addition Masterfile does not represent as wide a range of image partners as the big three. Given the number of images each of the big three represent, many of the image partner collections are effectively hidden no matter how good their keywording unless the subject the customer is looking for is very unique.
Are Medium An Small Supplers Destined To Lose Market Share?
If the big distributors were to implement this strategy is there anything medium and small agencies could do to keep from losing market share? Yes, but the transition process would not be easy.
Smaller and niche agencies and individual suppliers need a new central database for their images. Customers want one central place to go to look for professional images. But, each supplier also needs to be able to control the price charged for his/her work.
One key to the success of such a database is that once established suppliers would also need to pull their images out of Getty, Corbis and Alamy to keep them from undercutting the new database on price. This would be a very tough decision, but in the long run it may be the only way to get decent prices for their work.
The basic elements needed for such a site are all currently available, but no single site has everything. Alamy has some of what’s needed, but pricing is a big issue. PhotoShelter has inexpensive technology for a large database, but needs a better system for automatic direct sales and allowing each supplier to set its own prices. Masterfile has a working visual search system. Getty has shown that customers will use a site where prices vary greatly from image to image. (See “
Fixed Pricing vs. Use Pricing”) In addition the new site would need to embrace some of iStock’s features and be willing to license images at all price points, including microstock prices for certain uses.
The following are some of the other necessary characteristics of such a site.
1 - The database operator should offer pricing templates that suppliers can use for online licenses. Each supplier should have the ability to modify the template. (Current industry practices indicate that customers prefer simpler templates, but suppliers may also use their own more complex templates.)
2 – Each supplier may have multiple pricing templates for various groups of images in his collection.
3 – The system should be designed so most licenses are automatic online transactions, but negotiations and special deals would be allowed. Participating suppliers would agree by contract to pay the database operator a percentage of any license that is billed separately from the online, automatic system if the images was found by searching the system.
4 – Suppliers should be able to provide an automatic authorization for a customer to download one of their images from the site without a payment transaction.
5 – When customers want to negotiate a deals they must deal directly with the image creator or that creator’s primary agent.
6 – A system of credits that would enable volume customers to receive discounts should be included in the features offered, but every supplier would not be required to use this system. Customers would be given a choice of paying with dollars/pounds/euros or credits. A non-participating supplier would not offer the credit option for a particular image.
7 – Subscription deals would be allowed. Subscribers would be given a special password that would entitle them to unlimited access to any file in the collection to which they have subscribed.
8 – The site should allow search returns to be organized by downloads, newest images or an individual or group of suppliers. A visual search feature should also be included.
9 – Customers should be able to narrow their searches using features similar to those found on the Corbis site, i.e. RM, RF, Editorial, Documentary, Sports, Entertainment, Historical and Fine Art and the People and Image Attribute and Date categories. Those keywording images should be encouraged, but not required, to use the words that relate to these categories.
Costs:
1 – Suppliers would pay a monthly fee in order to participate on the database. The fee would be based on the number of images in the collection.
2 – The charge for the first pricing template would be included in the monthly charge. There would be an additional charge for each additional template.
3 – A percentage of all automatic sales would be retained by the database operator.
4 – Suppliers would also be required by contract to pay a percentage of negotiated sales.
If such a database were available, most image partners would probably want to pull their images from the big three to keep their prices from being undercut by the big distributors.