Keywording from an image users viewpoint

Whenever you are offering photos, you do so in order that other people can find and use them. To enable someone to find and use your photos, you add keywords and metadata. Does this preparation give you the desired results? That is a hard question for many people to answer. Keywording therefor must be made relevant and you should know which keywords are needed to help you achieve your goals. For this reason it is helpful to include the viewpoint of an image user in the process of keywording.

Crossing the floor

Now you are in a process of optimizing your kewords, it can be an eye-opening experience to cross the floor and for once take the position and viewpoint of an image user. It will help to get a better grasp of what keywording needs to achieve. Become a researcher who is exploring your own archive.

Let’s suggest that you look for an image of an egg. For an upcoming campaign you need a photo of an egg with certain requirements. As you might be aware, there are thousands of images with eggs in it (like here). Not all of those can be used for your campaign. You need to narrow down the search result and you do so by differentiating the search request. Adding keywords to be more precise is a good way to add more characteristics of the type of image you want to find and discover.  It’s not exactly brilliant to take just any picture of eggs.

Ask friends and colleagues

Keywording is a complex task, as people and the language(s) they use are constantly evolving and the use of words differs from person to person. Getting yourself in the position of a researcher already gives quite some insight. The next step could be to invite a few friends and colleagues to do your job. Send them one or two images and ask them to add 10 keywords to each image as they spontaneously popup in their minds. It will quickly show that not everyone uses the same words. The way we use language differs from person to person. To cope with this variety, the keyworder has to think beyond his own understanding and try to add enough differentiating words to cover as many search requests as possible.

There is this famous story, where six people are asked to describe the same room. Everyone writes his or her own impressions and you end up with six different stories. In comparison: When you are describing photos, try to imagine how other people would describe that same image. You then capture more relevant words reflecting different viewpoints.

Know your users

Focussed image libraries for clearly defined user groups logically need excellent understanding of the keyworder for both the user groups and the topics involved. Even “regular” images put into a different context might bring across a different message. Because of that one might need to adapt its keywords. For companies there are frequently additional words needed, to reflect its own marketing style as well as its own special needs and deeds. This results in extra keywords, acronyms or even word groups. Whenever you are working for a specific circle of interest or group fo clients needs to understand the context and wording of that specific group.

Apart from the appropriate technical expressions one also can think of abbreviations, usages (marketing, archive), events (Meeting Day 2014) or purpose (Winter campaign 2018, Easter 2019, Special Day 2022). For all these extra words, usually unknown for external keyworders, it is helpful if someone from within the company does or enhances the keywording or at least gives a thorough briefing about the expected keywords. This also is a good starting point for a fruitful cooperation between a keywording consultant and a company, museum, or archive.

Keywording should lead to results

Languages are alive, just as the people speaking them. There will never be and can never be a 100% reliable and comprehensive keywording. Over time even the interpretation of images might change and “modern” will turn “retro” if you just wait long enough. Thus the best you can get in keywording is an approach. We can try to get as near as possible to a description of an image. The main question is not about perfection, but rather about the results you can achieve with a proper keywording. A result-oriented thinking should help as a general guide-line for any keyworder.

The real goal of keywording is the ability to locate, differentiate and retrieve the proper image for your goal. Keywording is made for that process. If the user finds the appropriate image, the goal is reached. If he doesn’t find the image (though it would be available), something important might have been overlooked. The interaction between keyworder and image user therefor are paramount. A regular feedback from users of the image database can give valuable input. A successful search shows the quality of keywording.

If it is possible to evaluate the search request for any website, this information too would give valuable feedback. Which words have been used? Do researchers use multiple words to narrow down the results? Which words gave zero results? Are these important (but missing) topics, or have the proper images not been found as the keywording was too limited? Questions like these help to understand and improve the keywording.

These considerations might help to define a keywording strategy and to put in place some keyword controlling. To evaluate search requests along with the available images and their keywording can improve the quality of keywording.

Image credits: All images come from the stockphotography collection as available at Kursiv photoagency.