As image databases get larger and larger, keywording becomes more and more important as photographers try to get their work high enough in the search-return-order for the images to be seen. Often creators must spend more time keywording than they spend taking pictures. In addition, image distributors are constantly coming up with new strategies that often necessitate going back and re-keywording images that have already been uploaded.
Alamy has recently updated its Alamy Image Manager. It is not getting good reviews from creators. See
here. Alamy has published a blog post about
best practices for tagging using the new system. Included in their recommendations are to “tag naturally.” What does that mean?
“Well, it means that if something is a multi-word single entity then you should tag it as such e.g. “Taylor Swift”, “South East Asian”, “Bible Belt”, “Red Leicester”, “The Rolling Stones”, or “Homer Simpson”. If you’re still using the older version of our image management tool you can enter a phrase by adding a comma or ” ” around the phrase. If you’re using the new Alamy Image Manager (rolling out across the site now) then you can just type in each word before submitting them as a phrase.
“What we wouldn’t advise is putting multiple entities/concepts into a single tag e.g. “Paris Eiffel Tower”. As things currently stand this will not have any particular effect on the search engine one way or the other but in future iterations of the search engine it may do. It would be more natural (and hence better as a longer term tag strategy) to go for “Paris”,”Eiffel Tower”.
“Our tagging system does not exclude constituent words of a tag from being searched for e.g. “Banff National Park” will still show up for “banff”,”national park” and “park” searches. This also means that “Rhinoceros Beetle” will show up in “Rhinoceros” searches, but the search engine tries to make sure that the beetle image does not appear too high up.”
“Avoid adding alphabetical lists of tags. Proximity of one tag compared to another used in a multi-word search can have an effect on where it appears in the sort order, which is why we’d recommend adding phrases/multi-word tags where appropriate.”
Supertags
Alamy’s system used to allow for 3 “Essential Keywords” for each image. These tags were weighted differently from regular tags giving the image preference in the search-return-order. With the new system Alamy is introducing “supertags”. Contributors can add up to 10 supertags per image. These tags should be what the photographer believes are the words most customers might use when looking for an image of this type. The search engine places extra priority to supertags.
For help on adding tags and supertags in the Alamy Image Manager watch their
overview help video or read through the relevant section (page 16) in the
PDF instructions.
Problems
There are several problems for contributors. First, it takes a lot of time and careful thought. If the right tags aren’t on the image no one will see it.
Second, contributors have no idea how frequently customers use a particular word to search for images. A contributor may include a word as a supertag that he thinks is very appropriate, but, in fact, no customer has ever used that word to search for such an image. Thus the word is wasted. If photographers could access a list of word customers use, and the frequency of such use when they actually purchase an image, it would be very helpful in determining what words to use as supertags.
Photographers are visual people. They need to be able to see what images are actually selling when a customer uses a particular search term. What’s in demand and what isn’t. Broad general categories of what’s in demand are meaningless because there are such a broad range of approaches to the subjects in these categories. If, when a photographer is trying to determine what word to use as a supertag, the photographer could do an image search to see the specific images that were purchased when the word was used by customers in the past, it would be of tremendous help. Such a database could also help photographers produce more images that are on target with what customers want to buy.
Fourth, every distributor has a different search algorithm. Photographers working with several different distributors often need to adjust their keywording for each one. There may be no solution for this problem. But it should be recognized that the extra work involved discourages photographers from working with multi-distributors. Since there are very few cases where a photographer receives enough revenue from single distributor generates to justify continued production, it also discourages photographers from producing more imagery.
In an effort to try to make it easier for customers to find the right image it may be unwise to load more work on the backs of the photographers.