Marb Posted August 20, 2018 Share Posted August 20, 2018 I am asking this because as I strive to find as many relevant keywords as possible, I am finding some images coming up on the first page, first image with as little as 3 keywords. How does that happen. For instance, Marks and Spencer shop signs, the third image on the first page has 7 keywords. The seventh has just 3 keywords. So how do these images with such few keywords make it to the highest ranking on the search engine here ? Also, there is no difference in order if I select "Creative" or "relevant". Just want to understand how this site works so I can give myself a decent chance of being seen but it makes no sense so far. Link to comment Share on other sites More sharing options...
spacecadet Posted August 20, 2018 Share Posted August 20, 2018 The difference is that the images you cite have a high ranking earned from sales. There's not much you can do about that until you have them. Link to comment Share on other sites More sharing options...
Marb Posted August 20, 2018 Author Share Posted August 20, 2018 Just now, spacecadet said: The difference is that the images you cite have a high ranking earned from sales. There's not much you can do about that until you have them. Ok, but how did they get the sales in the first place with so few keywords ? Link to comment Share on other sites More sharing options...
Doc Posted August 20, 2018 Share Posted August 20, 2018 I think some customers will look for very specific images like an M&S shop sign - so how many keywords do you need? Kumar Link to comment Share on other sites More sharing options...
Cryptoprocta Posted August 20, 2018 Share Posted August 20, 2018 26 minutes ago, Marb said: Ok, but how did they get the sales in the first place with so few keywords ? Remember that the caption is also searchable. I've seen a photo with 0 keywords top of a search of a few hundred. Also consider that the fewer the keywords, the less chance there is of false juxtapositions - e.g. over the w/e three of my files showed up on a search for "George Conway" All three were taken in 'George' Square: two had the surname Conway and one had a news caption which described him as the sports partner of 'Robert Conway', so Conway wasn't in the keywords. False juxtapositions affect your click through rate. Link to comment Share on other sites More sharing options...
Marb Posted August 20, 2018 Author Share Posted August 20, 2018 4 minutes ago, Doc said: I think some customers will look for very specific images like an M&S shop sign - so how many keywords do you need? Kumar Still doesn't make sense. Link to comment Share on other sites More sharing options...
Joseph Clemson Posted August 20, 2018 Share Posted August 20, 2018 Images with many keywords will tend to show up in searches where the buyer is not actively looking for an image like that. This is especially the case if some keywords are only marginally relevant to the image in question. Every time an image shows up in a search and is not zoomed, it counts against your CTR (click through rate). Your CTR is one element of the secret formula which determines how your image in a returned set of photos in a search is ranked. The lower the rank, the further down the search. Images with only a few keywords (and being well captioned), will tend to show up in fewer searches BUT they will be highly relevant to those searches and are more likely to be zoomed and/or sold. The pseudonym with which they are associated will, over time, gain an higher rank, putting their other pictures higher in the search results too. This is why I (and others) argue that using the maximum number of keywords (and obsessing with getting the discoverability bar to green, as some still do), is in actual fact counter-productive. Link to comment Share on other sites More sharing options...
Marb Posted August 20, 2018 Author Share Posted August 20, 2018 1 minute ago, Joseph Clemson said: Images with many keywords will tend to show up in searches where the buyer is not actively looking for an image like that. This is especially the case if some keywords are only marginally relevant to the image in question. Every time an image shows up in a search and is not zoomed, it counts against your CTR (click through rate). Your CTR is one element of the secret formula which determines how your image in a returned set of photos in a search is ranked. The lower the rank, the further down the search. Images with only a few keywords (and being well captioned), will tend to show up in fewer searches BUT they will be highly relevant to those searches and are more likely to be zoomed and/or sold. The pseudonym with which they are associated will, over time, gain an higher rank, putting their other pictures higher in the search results too. This is why I (and others) argue that using the maximum number of keywords (and obsessing with getting the discoverability bar to green, as some still do), is in actual fact counter-productive. Thanks Joseph, that makes some sense at least. Link to comment Share on other sites More sharing options...
Cryptoprocta Posted August 20, 2018 Share Posted August 20, 2018 3 minutes ago, Marb said: Still doesn't make sense. Maybe, and I haven't a clue, the fewer the keywords, the higher each one is ranked? That actually would make some sense, as the fewer keywords, the less likelihood there is of spam. Link to comment Share on other sites More sharing options...
spacecadet Posted August 20, 2018 Share Posted August 20, 2018 As usual with rank, the answer is "all of the above". All things being equal a highly ranked image will appear higher in searches. But tight keywording is also rewarded with a better CTR, which also influences rank. Link to comment Share on other sites More sharing options...
Starsphinx Posted August 20, 2018 Share Posted August 20, 2018 4 hours ago, Cryptoprocta said: Remember that the caption is also searchable. I've seen a photo with 0 keywords top of a search of a few hundred. Also consider that the fewer the keywords, the less chance there is of false juxtapositions - e.g. over the w/e three of my files showed up on a search for "George Conway" All three were taken in 'George' Square: two had the surname Conway and one had a news caption which described him as the sports partner of 'Robert Conway', so Conway wasn't in the keywords. False juxtapositions affect your click through rate. This may be a dumb question but I am still learning the ropes of the image manager bit - how do you discover what searches your files have shown up in? My images are showing as having been searched but I have no zooms - how do I find out what people are looking for when they find my images so I can make any necessary changes to keywords etc Link to comment Share on other sites More sharing options...
Joseph Clemson Posted August 20, 2018 Share Posted August 20, 2018 50 minutes ago, Starsphinx said: This may be a dumb question but I am still learning the ropes of the image manager bit - how do you discover what searches your files have shown up in? My images are showing as having been searched but I have no zooms - how do I find out what people are looking for when they find my images so I can make any necessary changes to keywords etc You needs to go to your dashboard page, Alamy Measures, Your Images. The summary of pseudonyms it shows defaults to the last 30 days. If you click on a pseudonym it will show the which of your images have been returned by the search term specified. You can adjust the date range - anything from a single day to the whole of the last 12 months. No search data is available to us older than 12 months ago. Link to comment Share on other sites More sharing options...
Starsphinx Posted August 20, 2018 Share Posted August 20, 2018 5 minutes ago, Joseph Clemson said: You needs to go to your dashboard page, Alamy Measures, Your Images. The summary of pseudonyms it shows defaults to the last 30 days. If you click on a pseudonym it will show the which of your images have been returned by the search term specified. You can adjust the date range - anything from a single day to the whole of the last 12 months. No search data is available to us older than 12 months ago. Thank you I will see what I can find. Link to comment Share on other sites More sharing options...
M.Chapman Posted August 20, 2018 Share Posted August 20, 2018 10 hours ago, Joseph Clemson said: Images with many keywords will tend to show up in searches where the buyer is not actively looking for an image like that. This is especially the case if some keywords are only marginally relevant to the image in question. Every time an image shows up in a search and is not zoomed, it counts against your CTR (click through rate). Your CTR is one element of the secret formula which determines how your image in a returned set of photos in a search is ranked. The lower the rank, the further down the search. Images with only a few keywords (and being well captioned), will tend to show up in fewer searches BUT they will be highly relevant to those searches and are more likely to be zoomed and/or sold. The pseudonym with which they are associated will, over time, gain an higher rank, putting their other pictures higher in the search results too. This is why I (and others) argue that using the maximum number of keywords (and obsessing with getting the discoverability bar to green, as some still do), is in actual fact counter-productive. Also... any images that have been zoomed previously using the same search terms as the current customer's search are given a higher placement. Mark Link to comment Share on other sites More sharing options...
Bill Kuta Posted August 20, 2018 Share Posted August 20, 2018 Aren't rankings by photographer, not by individual image? Link to comment Share on other sites More sharing options...
M.Chapman Posted August 20, 2018 Share Posted August 20, 2018 11 minutes ago, Bill Kuta said: Aren't rankings by photographer, not by individual image? Bit of both. If you keep a record of your zooms (and the search term used) it's easy to see the "zoom" effect. If you have several images of the same subject, but only one has been zoomed then that image will appear first (amongst your images) if the search is repeated. A previous zoom seems more powerful than whether the search terms appear in the tags, supertags and/or caption in determining image position. Mark Link to comment Share on other sites More sharing options...
Ed Rooney Posted August 20, 2018 Share Posted August 20, 2018 12 hours ago, Joseph Clemson said: Images with many keywords will tend to show up in searches where the buyer is not actively looking for an image like that. This is especially the case if some keywords are only marginally relevant to the image in question. Every time an image shows up in a search and is not zoomed, it counts against your CTR (click through rate). Your CTR is one element of the secret formula which determines how your image in a returned set of photos in a search is ranked. The lower the rank, the further down the search. Images with only a few keywords (and being well captioned), will tend to show up in fewer searches BUT they will be highly relevant to those searches and are more likely to be zoomed and/or sold. The pseudonym with which they are associated will, over time, gain an higher rank, putting their other pictures higher in the search results too. This is why I (and others) argue that using the maximum number of keywords (and obsessing with getting the discoverability bar to green, as some still do), is in actual fact counter-productive. That is very well stated, Joseph. What I don't understand is why Alamy suggests we use 50 tags. I do not nor could not put up 50 relevant tags on any of my images. Relevant. That's what good tagging is about. Link to comment Share on other sites More sharing options...
Cryptoprocta Posted August 20, 2018 Share Posted August 20, 2018 Zooming (or 'something') seems to trump sales! I have a series of 38 images (all different but with one common generic term). Ten of these have sold, two of them twice. But on a sitewide search, the one which comes #2 of 2,355 hasn't sold (and hasn't been zoomed in the past calender year, don't know about before that). Link to comment Share on other sites More sharing options...
gvallee Posted August 20, 2018 Share Posted August 20, 2018 1 hour ago, Ed Rooney said: That is very well stated, Joseph. What I don't understand is why Alamy suggests we use 50 tags. I do not nor could not put up 50 relevant tags on any of my images. Relevant. That's what good tagging is about. I found that with plurals, stemming, phrases, variations with/without hyphen, abbreviations (Queensland/QLD) it's easy to get to 50 relevant tags. Especially if there are several concepts to the image to describe. Let's take the example of one person watching a sunset with clouds on a beach with palmtrees. Concept 1: landscape Landscape, view, scenic, picturesque, vista, Concept 2: mood Ethereal, Moody, Atmospheric, Sunset, Concept 3: coast Coastline, Coast, Beach, Beaches, Sandy Beach, Tropical Beach, Exotic Beach, Palmtrees, Palm Trees Concept 4: clouds Clouds, Cloudscape, Cloud Formation, Cloud Formations Concept 5: People 1 Person, One Person, People, Someone, Female/Male, Caucasian, Holidays, Vacation, Watching Sunset Concept 6: location Cairns, Queensland, QLD, Far North Queensland, FNQ, Australia You get the idea. Gen Link to comment Share on other sites More sharing options...
Niels Quist Posted August 21, 2018 Share Posted August 21, 2018 10 hours ago, gvallee said: I found that with plurals, stemming, phrases, variations with/without hyphen, abbreviations (Queensland/QLD) it's easy to get to 50 relevant tags. Especially if there are several concepts to the image to describe. - add to this American and British words and spelling - and sometimes common misspellings. Link to comment Share on other sites More sharing options...
gvallee Posted August 21, 2018 Share Posted August 21, 2018 31 minutes ago, Niels Quist said: - add to this American and British words and spelling - and sometimes common misspellings. Absolutely. I do that too. Quite tedious. Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.