MicrostockGroup
Agency Based Discussion => Alamy.com => Topic started by: Brasilnut on July 28, 2017, 11:12
-
I'm making a huge effort to turn almost 3000 images from poor discoverability to good without spamming. Not easy.
Anybody else had more success after doing so?
I'm using the following keywording tool which seems excellent:
https://www.keyword.io/tool/stock-photo-research (https://www.keyword.io/tool/stock-photo-research)
-
Word on the forum is that it makes no difference. I certainly won't waste my time imagining possible tangential keywords. It's a stupid invitation to spam. Because of the way their search works, I get enough false hits there already.
-
Word on the forum is that it makes no difference. I certainly won't waste my time imagining possible tangential keywords. It's a stupid invitation to spam. Because of the way their search works, I get enough false hits there already.
I agree, it is really only a waste of time.
The only way to get the "high discoverability" is to put more that 45 keywords I think, including 10 supertags.
But most of the images (mine) don't need 45 keywords.
-
Word on the forum is that it makes no difference. I certainly won't waste my time imagining possible tangential keywords. It's a stupid invitation to spam. Because of the way their search works, I get enough false hits there already.
Agree also by adding "tangential" keywords you will probably reduce the ctr which might affect your ranking. This and their half hearted attempt at categories seem rather daft.
-
I'm just slowly doing the supertags. Only 2 images optimised out of 2800.
-
I'm just slowly doing the supertags. Only 2 images optimised out of 2800.
I'm not fully sure how much effect supertags are having.
Make sure you keep checking regularly, it keeps changing.
Certainly a few months back, 'supering' a tag made a files fall back quite a bit in a search for the supertag, judged by before and after checks. I don't think it's so bad now.
-
I have up on it all a few months ago. I just don't make enough there to justify the time to get uploads supertaged.
-
Word on the forum is that it makes no difference. I certainly won't waste my time imagining possible tangential keywords. It's a stupid invitation to spam. Because of the way their search works, I get enough false hits there already.
Is that five yes to your five answers. I think so. :) It's another carrot on a stick mission, from some stock agency.
-
Certainly time consuming but if it improves sales by even 10% I think it's worth it. I did key word some in a rush the first time round.
-
Certainly time consuming but if it improves sales by even 10% I think it's worth it. I did key word some in a rush the first time round.
But if even more spammy keywording leads to buyers leaving ...
As they don't have any auto-translation, it might be worth the effort to translate the main subject keywords - but it would be difficult to guess which languages would work best, unless you have an obvious bias in sales location. Certainly adding main keywords in the language of the location of images could be helpful. Though it could also clog up searches where the same word has a totally different meaning in different languages.
-
it would be difficult to guess which languages would work best, unless you have an obvious bias in sales location.
That's a great tip as many of my images are in Italy it makes sense translate them to Italian to appeal to the local market :)
-
My 400 images haven't had a sale since... let me check... oh yeah April. And before that, I had one in January. It's over. I don't even waste time checking for sales, let alone jumping through more keywording hoops.
-
Ha! It's a members game on Alamy. I didn't have my first sale until I had over 800 files there. Still, that one was good enough to keep me uploading. Nowadays 400, even 4000 is just a drop in the ocean. I
Everyone has to make their own decisions in this game - what works for some doesn't work for others
-
Hey,
thanks for mentioning keyword.io, I built that a while ago, happy to see it's helping!
I once made a quick test based on 30+ million stock photos in the database and there indeed was a strong correlation between the number of keywords and number of sales. In short: The more the better.
But:
1. Irrelevant keywords certainly don't help and might actually hurt by reducing your CTR
2. There might be a correlation between quality of the asset and number of keywords - more experienced and devoted artists may invest more in keywording.
So use tools like keyword.io to find keywords that you might have missed otherwise, but don't try to fill up all slots with tangential stuff :)
Also, make sure to optimise your title/description and use the most relevant keywords there, too. Basic search engine optimisation...
-
Hey,
thanks for mentioning keyword.io, I built that a while ago, happy to see it's helping!
I once made a quick test based on 30+ million stock photos in the database and there indeed was a strong correlation between the number of keywords and number of sales. In short: The more the better.
But:
1. Irrelevant keywords certainly don't help and might actually hurt by reducing your CTR
2. There might be a correlation between quality of the asset and number of keywords - more experienced and devoted artists may invest more in keywording.
So use tools like keyword.io to find keywords that you might have missed otherwise, but don't try to fill up all slots with tangential stuff :)
Also, make sure to optimise your title/description and use the most relevant keywords there, too. Basic search engine optimisation...
How does one do a "quick test" of 30 million images? And where would one get these sales figures?
-
Secret of trade ;)
But that's the data I gathered for http://www.keyword.io/tool/stock-photo-research (http://www.keyword.io/tool/stock-photo-research) into a big database, so it's a matter of minutes to run such queries on.
-
Secret of trade ;)
But that's the data I gathered for [url]http://www.keyword.io/tool/stock-photo-research[/url] ([url]http://www.keyword.io/tool/stock-photo-research[/url]) into a big database, so it's a matter of minutes to run such queries on.
But where do you get the sales figures? Top sites don't show or release these. Alamy doesn't release this data.
Personally I've found that good accurate specific keywords will get better sales than "more keywords" just for the sake of having more words. Imagine if you are a buyer and wanted a photo of something? What would you search for? That's the words to include, not obscure or random similar or minor details that aren't the main subject.
Too many people think there's some game to making sales, when the answer is honest and straight forward. Good accurate keywords work best. And yes I've tested my own images, against my own images of similar photos. Next time you upload something of a series, make ten with basic words, precise, make ten with many keywords, make ten more with all kinds of obscure and related words. Then wait a month and look at where they appear in the search.
I promise, less words, more accurate, will be ranked highest at most sites. This assumes that the repeating words are the same number of instances, because that has an effect at some sites.
-
Some agencies do publish those numbers in their APIs, although you may not see those on their websites.
First of all a big disclaimer: I am not selling photos myself so all I say is from looking at statistics and educated guesses.
I totally agree that quality > quantity and to be very clear: Keywording tools should be used to find more _relevant_ keywords, not just max out the amount of allowed keywords.
That doesn't mean, that you shouldn't add very *generic* keywords, too, though.
For the fun of it I just looked for the most frequent keywords in the database, see end of post.
12.5% of images have the keyword "background". Further analysis shows, that 79.566% of assets with that keyword have 0 sales while 80.076% of ALL assets have 0 sales. Now I don't want to do the math to test if that's scientifically significant, but I would make the case that generic keywords don't hurt your sales. Other top keywords rank around 79.5 - 80.5, too.
Which makes sense to me intuitively - you will never rank for "background" or if your CTR would be crap, but your average buyer won't search just for "background" but for "dense forest background" - where your picture "dense forest" picture will get its chance once you add the generic "background" keyword.
(Another shameless pluck: check out http://www.keyword.io/tool/fotolia-longtail-finder (http://www.keyword.io/tool/fotolia-longtail-finder) to get ideas of actual search phrases based on fotolias's autosuggest. It's far from being as good as google's though, so you might want to have a look at that, too...)
Here's some raw data of the 100 most frequent keywords:
[
{
"key": "background",
"doc_count": 4591388
},
{
"key": "white",
"doc_count": 3987768
},
{
"key": "nature",
"doc_count": 2484515
},
{
"key": "isolated",
"doc_count": 2435124
},
{
"key": "beautiful",
"doc_count": 2413848
},
{
"key": "people",
"doc_count": 2229395
},
{
"key": "design",
"doc_count": 2127786
},
{
"key": "illustration",
"doc_count": 1904740
},
{
"key": "blue",
"doc_count": 1888078
},
{
"key": "green",
"doc_count": 1886371
},
{
"key": "young",
"doc_count": 1874181
},
{
"key": "woman",
"doc_count": 1753721
},
{
"key": "red",
"doc_count": 1703703
},
{
"key": "person",
"doc_count": 1669872
},
{
"key": "female",
"doc_count": 1655572
},
{
"key": "summer",
"doc_count": 1603761
},
{
"key": "beauty",
"doc_count": 1568581
},
{
"key": "travel",
"doc_count": 1563713
},
{
"key": "color",
"doc_count": 1561565
},
{
"key": "business",
"doc_count": 1536681
},
{
"key": "on",
"doc_count": 1534946
},
{
"key": "food",
"doc_count": 1525440
},
{
"key": "symbol",
"doc_count": 1462259
},
{
"key": "vector",
"doc_count": 1444441
},
{
"key": "happy",
"doc_count": 1442463
},
{
"key": "concept",
"doc_count": 1404698
},
{
"key": "black",
"doc_count": 1355581
},
{
"key": "old",
"doc_count": 1328315
},
{
"key": "girl",
"doc_count": 1328268
},
{
"key": "of",
"doc_count": 1319744
},
{
"key": "sky",
"doc_count": 1319360
},
{
"key": "closeup",
"doc_count": 1314140
},
{
"key": "caucasian",
"doc_count": 1306044
},
{
"key": "in",
"doc_count": 1288970
},
{
"key": "healthy",
"doc_count": 1281133
},
{
"key": "art",
"doc_count": 1268722
},
{
"key": "man",
"doc_count": 1267496
},
{
"key": "natural",
"doc_count": 1241623
},
{
"key": "abstract",
"doc_count": 1221299
},
{
"key": "portrait",
"doc_count": 1202016
},
{
"key": "adult",
"doc_count": 1143252
},
{
"key": "fresh",
"doc_count": 1126741
},
{
"key": "water",
"doc_count": 1102621
},
{
"key": "with",
"doc_count": 1089499
},
{
"key": "style",
"doc_count": 1071221
},
{
"key": "architecture",
"doc_count": 1053418
},
{
"key": "decoration",
"doc_count": 1051273
},
{
"key": "yellow",
"doc_count": 1034675
},
{
"key": "male",
"doc_count": 1029181
},
{
"key": "landscape",
"doc_count": 1020224
},
{
"key": "sign",
"doc_count": 1015971
},
{
"key": "modern",
"doc_count": 1014797
},
{
"key": "outdoor",
"doc_count": 997496
},
{
"key": "vintage",
"doc_count": 993881
},
{
"key": "pattern",
"doc_count": 987256
},
{
"key": "cute",
"doc_count": 981877
},
{
"key": "lifestyle",
"doc_count": 980955
},
{
"key": "holiday",
"doc_count": 979195
},
{
"key": "and",
"doc_count": 977286
},
{
"key": "texture",
"doc_count": 968744
},
{
"key": "the",
"doc_count": 961560
},
{
"key": "colorful",
"doc_count": 959809
},
{
"key": "view",
"doc_count": 948446
},
{
"key": "light",
"doc_count": 941141
},
{
"key": "nobody",
"doc_count": 937180
},
{
"key": "fashion",
"doc_count": 921896
},
{
"key": "tourism",
"doc_count": 888435
},
{
"key": "icon",
"doc_count": 885558
},
{
"key": "outdoors",
"doc_count": 884011
},
{
"key": "brown",
"doc_count": 880305
},
{
"key": "building",
"doc_count": 875284
},
{
"key": "city",
"doc_count": 867419
},
{
"key": "graphic",
"doc_count": 864338
},
{
"key": "one",
"doc_count": 853655
},
{
"key": "bright",
"doc_count": 840991
},
{
"key": "retro",
"doc_count": 839586
},
{
"key": "attractive",
"doc_count": 834603
},
{
"key": "hand",
"doc_count": 833679
},
{
"key": "health",
"doc_count": 830408
},
{
"key": "technology",
"doc_count": 819292
},
{
"key": "day",
"doc_count": 817106
},
{
"key": "wooden",
"doc_count": 781709
},
{
"key": "object",
"doc_count": 777029
},
{
"key": "traditional",
"doc_count": 770831
},
{
"key": "smiling",
"doc_count": 769664
},
{
"key": "fun",
"doc_count": 768681
},
{
"key": "plant",
"doc_count": 767075
},
{
"key": "pretty",
"doc_count": 760217
},
{
"key": "wood",
"doc_count": 738521
},
{
"key": "face",
"doc_count": 725763
},
{
"key": "smile",
"doc_count": 716028
},
{
"key": "spring",
"doc_count": 713376
},
{
"key": "season",
"doc_count": 703530
},
{
"key": "tree",
"doc_count": 698072
},
{
"key": "sea",
"doc_count": 686473
},
{
"key": "paper",
"doc_count": 681462
},
{
"key": "europe",
"doc_count": 680923
},
{
"key": "culture",
"doc_count": 667678
},
{
"key": "animal",
"doc_count": 665415
},
{
"key": "home",
"doc_count": 659297
}
]
[Edited to put the data into a code section]
-
I have a hard time believing that an agency would make contributor sales figures available through a public API, even if they vetted and licensed users of that API and had them sign appropriate NDAs.
-
Several do. They are interested in making sales through resellers, so they give them quite a lot of data. On some you can even see the artist names...
[Edited]: To be fair, alamy.com is not part of the database, not sure if they do provide such numbers.
-
Then you've given me yet another reason to never have anything to do with those "agencies" again. Not that I needed one.
-
Several do. They are interested in making sales through resellers, so they give them quite a lot of data. On some you can even see the artist names...
[Edited]: To be fair, alamy.com is not part of the database, not sure if they do provide such numbers.
What agencies? SS, IS, FT, Alamy, 123RF, pond5, or irrelevant minor tossers.
-
It's a mix of major and minor agencies. Feel free to read through their tech documentation yourself, it's all openly available, but please understand that I won't give any details to make it easier for the competition to copy the tool ;)
-
I'm making a huge effort to turn almost 3000 images from poor discoverability to good without spamming.[/url]
Even less worth it now:
http://discussion.alamy.com/index.php?/topic/7987-download-packs (http://discussion.alamy.com/index.php?/topic/7987-download-packs)
-
It's a mix of major and minor agencies. Feel free to read through their tech documentation yourself, it's all openly available, but please understand that I won't give any details to make it easier for the competition to copy the tool ;)
I don't see many people who are buying your claims with no proof except you won't help the competition. What competition if no one cares about your empty tool. Show us something that means something.
-
I don't see many people who are buying your claims with no proof except you won't help the competition. What competition if no one cares about your empty tool. Show us something that means something.
Wow - what a friendly site we are today! Some one comes along with a free (and useful tool) to generate keywords and quite a lot of other interesting bits that you can use or ignore as you see fit and we hammer them into the ground.
Sorry, rjung, on behalf of some other members of the forum
Steve
-
I don't see many people who are buying your claims with no proof except you won't help the competition. What competition if no one cares about your empty tool. Show us something that means something.
Wow - what a friendly site we are today! Some one comes along with a free (and useful tool) to generate keywords and quite a lot of other interesting bits that you can use or ignore as you see fit and we hammer them into the ground.
Sorry, rjung, on behalf of some other members of the forum
Steve
+1
P.S. rjung, thanks for a nice tool, will certainly use it to improve my keywording skills.
-
I don't see many people who are buying your claims with no proof except you won't help the competition. What competition if no one cares about your empty tool. Show us something that means something.
As someone said recently on another thread: "Please can't someone find something else to bitch about besides things that aren't actually a problem?"
-
I don't see many people who are buying your claims with no proof except you won't help the competition. What competition if no one cares about your empty tool. Show us something that means something.
As someone said recently on another thread: "Please can't someone find something else to bitch about besides things that aren't actually a problem?"
https://english.stackexchange.com/questions/259479/a-word-for-someone-complaining-about-the-complaint
-
I don't see many people who are buying your claims with no proof except you won't help the competition. What competition if no one cares about your empty tool. Show us something that means something.
As someone said recently on another thread: "Please can't someone find something else to bitch about besides things that aren't actually a problem?"
https://english.stackexchange.com/questions/259479/a-word-for-someone-complaining-about-the-complaint
I was actually quoting the complainer's words back to him.
-
Thanks, glad to see the tool is useful to some! No worries, I don't feel hammered in the ground just yet ;)
I understand some people are wary about free stuff. I built it because it's fun, hopefully useful and a good way to promote another (free) service I built - dropstock.io (http://dropstock.io) - which maybe one day might make me a bit of money on the side. So much about ulterior motives.
If the tool is useful for you that's great. If it's not, I appreciate constructive feedback. If you feel you don't get any value out of the tool without knowing exactly which agencies data is used, that's too bad...
-
Thanks, glad to see the tool is useful to some! No worries, I don't feel hammered in the ground just yet ;)
I understand some people are wary about free stuff. I built it because it's fun, hopefully useful and a good way to promote another (free) service I built - dropstock.io ([url]http://dropstock.io[/url]) - which maybe one day might make me a bit of money on the side. So much about ulterior motives.
If the tool is useful for you that's great. If it's not, I appreciate constructive feedback. If you feel you don't get any value out of the tool without knowing exactly which agencies data is used, that's too bad...
I'm sure the tool will be useful, my only question was the ability to read our financial data and see how much we got for downloads, no doubts about the keyword tool. Others seem to read between the lines and get their undies all in a bunch.
Nice work, I'm sure I'll try it. Just was skeptical that major agencies has such lax security that you can read our earnings?
-
OK, didn't see that concern, my bad.
Those APIs only expose the amount of sales/downloads but there are no details about what kind of license, resolution, .... let alone amount of money earned.
That's not a matter of lax security but agencies putting in the effort of making this data available for third parties to sell more through resellers and integrations into big publishers software solutions. So it's in the interest of the artists, too.
Furthermore you can't see aggregated statistics per artist (you can just sum up sales on their assets you know about) or track over time (unless you want to do the delta between intervals). It's interesting to see the spread between artists' portfolios, some having a few dozen pictures, others have tens of thousands.
Apart from curiosity and vanity, I don't really see a point publishing such numbers, so I wont.
Sorry for the confusion, hope this helped clarifying :)
-
Thanks, glad to see the tool is useful to some! No worries, I don't feel hammered in the ground just yet ;)
I understand some people are wary about free stuff. I built it because it's fun, hopefully useful and a good way to promote another (free) service I built - dropstock.io ([url]http://dropstock.io[/url]) - which maybe one day might make me a bit of money on the side. So much about ulterior motives.
If the tool is useful for you that's great. If it's not, I appreciate constructive feedback. If you feel you don't get any value out of the tool without knowing exactly which agencies data is used, that's too bad...
Thank for your tool rjung. I used it for long, and found it very useful and easy to use.
As ex software developer I can appreciate the good job.