I see a lot of people with different takes on keywording (me included!) and I can't help but wonder if it isn't a self-fulfilling prophecy. Consider two very different approaches to keywording, along with my take on what the results would be:
1. Relevancy is the goal:
You create images with short titles, accurate descriptions with the title in the keywords and a minimal set of highly relevant keywords.
My take: by doing this you forgo the chance to get the occasional download based on a match of 3 or 4 keywords (someone searching on 'red apple with copy space'), but you are highly like to get found if someone looks for 1 or 2 main words ('paper').
2. Keyword matches are the goal:
You pay little attention to titles and description and make sure that you get all possible relevant keywords. Note that this doesn't mean spamming keywords, but you don't leave anything to chance in terms of a missed keyword. When someone does a complicated search you are sure to be found 'american eagle flying fish'.
This isn't about which method is better, or other ways of keywording (though I clearly favor high relevancy), but if you consider the two I think they will both 'succeed'. In the first case, you are highly likely to get found if someone searches on the title of your image, but probably not via many other ways. This reinforces the belief that the title is important. Alternatively, in the 2nd case you will be more likely to get found via a 3 or 4 word keyword match - reinforcing the belief that you need lots of keywords. To go even further, I suspect that being exclusive plays into it too. If you are exclusive (I am), then you will be placed higher in the results and do better in a single term keyword search with its thousands of matches. As a non exclusive that doesn't make it to the first page, you probably will do better with more keywords...
What are your thoughts? Do your images get found with lots of different keywords or only a few, are your sales all n/a type or mostly keywords, etc.?
My advice on keywording is to read what others have to say (especially the admins), experiment and see what works for your images - and look at the person giving advice and see if they are exclusive or not, and how many sales they have...
In the end, I do think it is a bit of a self-fulfilling prophecy - even if you get things totally wrong you will eventually get sales if you have enough images. And the feedback from those sales will tend to reinforce your beliefs about how buyers search for images. Don't be afraid to take some chances with your images and keyword some in different ways and see how they do - just be sure to give it enough time to work through the system.
In the meantime, I'm still waiting for the day that someone buys this image with a keyword other than 'baseball':
Your comment must be written in English.
We value all opinions and we will not censor or delete comments unless they come from fake accounts or contain spam, threats, false facts or vulgarity.
Bradcalkins
Agreed - almost all descriptive. One notable exception for me is 'future'...
Tan510jomast
Way to go Brad, congrats.
There you go again, another sales with only "descrriptive" keywords.
I did the same double take of my download list, and nowhere have I seen any solid evidence that conceptual keywords dominate the buyers' search.
Instead, I see more literal keywords , bicycle, boy, family, panel, etc.. You know what I mean.
Bradcalkins
Indeed - it is hard to know if the quality gets the sales, or the concise keywords! In the end I have to think keywords trump quality - if no one finds your images quality takes a lot longer to be discovered. Both are a kock out punch :)
Update: I have sold my baseball with a three keyword search - 'well used baseball'. That highlights something I struggle with: 'well' on its own is not a good keyword for a shot of a baseball, yet totally relevant in this context, and used by a buyer to find it.
Tan510jomast
Yes, I agree keywords can make or break your images. Many lesser images tend to sell , to the amazement of the creator . It's truly a dilemma to go conceptual or descriptive.
Still, I believe it's a lot of luck, and depends on how you call it.
For example, today I read about one of my own , a lady from my home province who celebrated her long career of super success as what I named her "Wonder Woman of micro stock", because I had earlier wrote her an email to congratulate her on her millionth downloads.
Checking her habit of keywording, I noticed she uses very little , mostly direct and descriptive. How many? no more than 10 .
From this, she proves spamming really sells little.
BUT OF COURSE, her images are incredible too, so it's both quality in photograph and good simple relevant keywords. ie 7-10 .
Back to the drawing board for me, I have lots of keywording adjustment to do,lol... I want to succeed like my new found micro stock idol Wonder Woman.
Anhong
Good article. Thanks!
Creativei
Yeah brad your assumption is right, the key words in sold images are sorted by alphabetical order.
Bradcalkins
"Oh just went through my sales report and found an interesting pattern of keywords used, most I guess around 55 percent of keywords used for search is in reverse order,
Eg:
Instead of search for Dubai Shopping Festival, buyer searched festival dubai shopping, instead of word map and vector, buyer used map word, and instead of Dubai Media City, buyer searched city dubai media"
Interesting - my assumption based on my used keywords has been that the keywords are just shown in alphabetical order, not according to the actual search. Does that agree with the order you see? They also strip out minor words like 'of', 'the', 'a', etc.
Jmphoto
Great blog Brad. I just looked through most of my recent downloads and most of them seem to have found the image through a word used in the title which are usually the most relevant words anyway!
Creativei
Oh just went through my sales report and found an interesting pattern of keywords used, most I guess around 55 percent of keywords used for search is in reverse order,
Eg:
Instead of search for Dubai Shopping Festival, buyer searched festival dubai shopping, instead of word map and vector, buyer used map word, and instead of Dubai Media City, buyer searched city dubai media.
Creativei
As a buyers I don't go beyond page 6, If i don't find in the first three page, i might search until page 6, beyond that never, well in other sites we use lots of and when we search (for Eg pretty and girl and white and asian)
Well most of my sales have more than two keywords. And finally brad a good blog.
Preckas
Very nice. :)
Antoinettew
Thanks Brad. My sales with 1 or 2 keywords are 50%. Allthough it takes time I think its fun to get the most out of keywording. If someone could correct me: be my guest. I would like to hear if I could do better.
Bradcalkins
I'm not producing images with a huge concept behind, so I can stick to behind descriptive I think! (a bit like your baseball image needs probably less keywords than your clean air concept)
I would agree wtih that - the simpler the image the fewer keywords are probably required. I often struggle with a complication or unique shot - you are much less likely to get a single keyword match, so it probably makes sense to flush out the keywords as much as possible...
Bradcalkins
I am not very familiar with keyword search results but please anyone who knows how the system work correct me. If many images pops up after a keyword or series of keywords input how much patience will the searcher and prospective buyer go to the extend of selecting an image he or she wants? Like a result of say maybe 10,000 or more images appearing? Will the older images be left behind overlooked if they were far back in the accepted queue?
Hard for non-buyers to know, but whenever I've searched for an image I probably stop after 5 or so pages and refine my search rather than endlessly looking. See my blog on searching to see how someone might use the many options DT offers: Using DT Search Tools
Meryll
Hi Brad,
my experience is that my photos are sold with one , more often with two or sometime three keywords. Usually one of the keywords is used right in the title at least. Knowing this I prefer to title my pictures with more words than one. I use rather more keywords to match various buyer's searching possibility. I have also N/A sales ( I would say 40-50%) alhough I have no collections yet.
Heywoody
No so much to go on yet, quite a few N/As, some dumb luck to be picked from many thousands but, where key words have worked, they have been the basic obvious ones used to descripe the picture.
Rosedarc
Hi Brad, thanks for this blog that triggers some thinking!
In my case, I go for short titles and try to be accurate and not overdo it in the keywords part - stick to the selling point of the picture and not go into irrelevant details (easier said than done of course). About 35% of my pictures were sold under N/A. I don't exclude, however, that they were initially found with a keyword, as I have this theory that quite a few N/A sales come from the buyers using the light box storing the images there until they've made the final call on the image (for a small portfolio that it). All other images were found with one, two or max three keywords (only 2 had 4 keywords), and the keywords were simple words describing my images.
Then again I'm not producing images with a huge concept behind, so I can stick to behind descriptive I think! (a bit like your baseball image needs probably less keywords than your clean air concept)
I think ultimately that each case scenario is different, exclusive / non exclusive, size of the portfolio, type of images in the portfolio...
Trottola
Very interesting blog. I think that a little luck can always help in selling too ;). thanks for sharing.
Egomezta
Thanks for sharing.... Good to know..... : )
Unteroffizier
I am not very familiar with keyword search results but please anyone who knows how the system work correct me. If many images pops up after a keyword or series of keywords input how much patience will the searcher and prospective buyer go to the extend of selecting an image he or she wants? Like a result of say maybe 10,000 or more images appearing? Will the older images be left behind overlooked if they were far back in the accepted queue?