Recently on the Linkedin
Stock Photography blog Valerie Henschel asked, “When do you cull older non-selling images from your archive?”
It is certainly something to think about. If customers are forced to go through a lot of outdated, mediocre or totally irrelevant images in order to find something that really fits their needs – and hopefully the best of that subject matter available in the collection – they are likely to give up and go elsewhere. As the choices of almost any photographic subject expand exponentially, this is becoming a bigger and bigger problem for buyers.
But, David Davies pointed out, “I had pictures which are between 20 & 30 years old licensed for the first time last year. I leave my old images with my stock library, you never know what is going to sell!”
Diego Lezama said, “I am a Corbis photographer and asked myself the same question, but the surprise came when Corbis sold a photo I took 20 years ago for Paramount Films (WWZombies). It was a biggggg sale and I never imagined this could happen! So, my thinking is, the image must be there just waiting for the right client. “
Steve Bisgrove said, “I often sell stock images that I took on film over twenty years ago. I don't see any reason to remove an image just because it's old. If the image is relevant to the buyers needs then it will sell. As for expanding the volume of images... Image libraries add thousands of photos to their collections every week. If we all removed our photographs that we thought were too old I think it would just be a drop in the ocean compared to the amount that would be left on-line.”
Diane Macdonald said, “A lot of my older stock images still sell quite well and some will surprisingly pop up and sell for the first time. My best selling image ever was photographed 20 years ago on 4x5 transparency film and actually copied into 35mm format on a lightbox! It used to sell many times a month, and still licenses at least once a month every month, but for a lot less money! And get this - it's an image of a city skyline!!”
Eli Vega said, “As long as an image does not look outdated, there is no such thing as an old image. Bottom line: It's not when an image was created that dates it; it's the ‘what.’"
Alfonso Guiterrez of AGEfotostock (www.agefotostock.com) echoed the “what” and said, “Almost all images have possibilities irrespective of when they were shot if good information is attached to them, something that not that many photographers realize. A Cincinnati skyline is different today (I believe) than it was 20 years ago, and while a 20 years old image may not serve current needs to show the present aspect of the city it may be of value for historical reasons provided that the date when it was taken is included in the caption of the image, something that photographers are resistant to include.
“I have and image that shows the ‘communicating vessels’ (name given to a set of containers containing a homogeneous fluid: when the liquid settles, it balances out to the same level in all of the containers regardless of the shape and volume of the containers) that I have sold close to 600 times. I took the image in 35 mm. Ektachrome in 1971, scanned as digital in 1998, and I still sell it almost every month, I have been tempted to repeat it but I broke the crystal device that I built to show the principle in 2003 and I´m too lazy to have it built again.”
Is There A Solution To The Too Many Images Problem?
On thing seems clear. Almost no one culls the images in their collections. Given the time it takes to review and decide which images to remove, it is easier to just keep adding images. But as image databases get larger and larger they often become less efficient for the customer. The more images to choose from, the harder it is for the customer to find the right image at the right price for his project.
Alfonso continued, “If we are talking about agencies here, I shall say that (a) adding more filters to the search (many agencies don´t have enough of them), (b) using keywording disambiguation (very few agencies do because it is complex and needs a controlled dictionary that has to be maintained, (c) filter by color, (d) filtering by similar (and exact match) and to top it all, (e) filter by date. The last, (e), depends on information supplied by photographers and this is a double-edged sword, clients like to ask when the image was taken and when they know the date they feel, in some cases, that the image is too old, although it shows what they wanted, and move forward to find a more recent one, something that suggests that dates have to be administered with caution. However a "by date" filter helps but again the agency has to decide whether to classify by the date the image was taken (that probably the photographer didn't supply) or the date that image was received by the agency.”
Database operators keep harping on keywording but for many of the subjects in highest demand keywording alone cannot adequately describe an image.
Suppose a customer needs a picture of a “woman” in an “office” environment using a “cellphone.” Search for those three words and Alamy gives you 4,386 images, Shutterstock 6,560 and Getty 5,617. No customer will take the time to go through more than a few hundred of the images on any of these sites. In addition, take a look at what these sites offer when you do this search and you’ll find a significant number of the images that are only marginally appropriate to this request. But, assuming all the images were right on target for these three words, so much of what the customers is looking for depends on the
look of the environment when deciding what is the best image. Adding keywords that accurately describe the environment might help, but only if the customer thinks to use the same keywords as the photographer attached to the image.
This problem occurs with many subjects. Sometimes more keywords can result in too many irrelevant image being included in certain searches. On the other hand limiting keywords to just the core elements often results in appropriate images not being found because the customer approached the search from a different angle.
What is needed to improve search is human editing, but that takes time and costs money
unless we can take advantage of the editing our customers are doing for us.
Tomorrow, I’ll offer a solution that could make it easier for our customers to more quickly find the best images for their projects at prices they can afford.