pancakes

MicrostockGroup Sponsors


Author Topic: Alamy Measures - Someone help me understand?  (Read 2168 times)

0 Members and 1 Guest are viewing this topic.

Uncle Pete

  • Great Place by a Great Lake - My Home Port
« on: April 15, 2022, 12:21 »
+1
I see a few people who say you can learn from Alamy keywords words used to find images and make sales. I don't see how? Can someone explain? Do you mean my own images or all of Alamy or what? And HOW?


Last Updated on : 15 Apr 2022 11:28
Results per page: Page 1 of 16161
Search Term  UCO Sales Zooms Views CTR (%)
DOLLY PARTON    17   88   179   16,975   1.05
putin    63   68   127   39,009   0.33
elon musk    48   59   147   30,810   0.48
marbella spain    1   49   63   2,100   3.00
Mariupol    72   35   126   42,448   0.30
Bruce Willis    30   35   66   23,185   0.28
vladimir putin    27   31   59   25,724   0.23
rishi sunak    61   29   65   20,826   0.31
boris johnson    62   27   91   29,838   0.30
street in lisbon    1   26   44   4,500   0.98
ukraine refugees    41   26   77   15,345   0.50
roman abramovich    29   26   64   14,651   0.44
pep guardiola    21   24   39   11,305   0.34
donald trump    20   22   71   11,519   0.62
zelenskyy    17   21   27   10,022   0.27
jim carrey    5   20   26   5,134   0.51
joust    1   20   41   980   4.18
ten hag    13   20   16   6,063   0.26
UKRAINIAN REFUGEES    23   20   32   5,031   0.64
ferragudo    1   19   31   2,300   1.35

Those are sold images. I don't have one with any of those words. I can go way down the list and I don't have anything that searched or sold recently.

Or am I supposed to use the data for MY images only?



First four are a match, the rest except winners podium have nothing to do with any of my images. False match. What do I learn from this?


zeljkok

  • Non Linear Existence
« Reply #1 on: April 15, 2022, 15:21 »
0
It is very simple.  If you look whole Alamy you will get overall trends, but there will be lots of data.  I generally look at my own more.  Consider attached file

Search Term tells you keywords customers were searching for. Important to note: Not ALL customers, only registered ones
Your Views tells you how many (my) images were returned for this search term
Sessions not really useful, you can ignore
Your Zooms Very important!  Tells you that how many times customer actually clicked (zoomed) my image.  If >0, it is good - means it picked their interest. This is potential sale;  sometimes sales get reported some time after the zoom. 
Your Sales Self explanatory
Your CTR - Click Through Rate - number of zooms / divided by number of views x 100.  CTR brings up  Alamy Rank - the more zooms, the higher the rank - means images are sorted higher in customer searches
Total Views Total number of images, including mine, that were returned for given Search Term.  Basically tells how many images I was competing against for that search term
Total CTR Total CTR for given search term.  If 0, it means customer really didn't find anything interesting


So for instance in my example, for search term "Kilimanjaro Summit" I had 2 images returned,  1 was zoomed, none sold, click-through was 50 (1 / 2 x 100), Total 100 images were returned across whole Alamy, of which 4 were zoomed with CTR of 4.  So I did fairly well,  customer zoomed at 4 images out of 100, and 1 was mine.

The most useful column by far is Search Term. You see what customers are looking for.  Then you think "Wait, I have such photo - why was it not returned in search?" You go to Alamy Image Manager, find the photo and see that maybe you forgot some of words in Search Term. Of toggle it to super keyword, Etc.

Or, on the other hand, you can identify "False Positives" - your image was returned but it is not at all representative for search term.  Naturally, it will not be zoomed, and thus bring down CTR and Alamy Rank. So you go to Image Manager and remove or at least adjust again the keywords.

« Last Edit: April 15, 2022, 15:54 by zeljkok »

« Reply #2 on: April 15, 2022, 15:50 »
+1
I too wonder why some really work the AoA data.  It's too much work for me.

My secret tweak is as follows:  I have two pseudos.  (In prior days psuedo rank was really important for search placement.  Still seems to be although it doesn't vary much anymore.)
My top pseudo consists of sold or zoomed images.  It has a high resulting rank.  So when an image is sold or is zoomed, I change the pseudo for the image.
This is also a fair bit of work, but keeping it up daily when I check sales, it's easy.

Uncle Pete

  • Great Place by a Great Lake - My Home Port
« Reply #3 on: April 17, 2022, 21:31 »
+1
How does looking at all this data, predict anything for the future or help me create anything that might get more downloads.

I can understand ignoring ALL as none of those have anything to do with anything I shoot. Looking at sales is history not the future. Most of the time it's subjects and places I've never been and never will be. Some are specific subjects I couldn't shoot if I wanted to.

I supposed I should ignore that and concentrate on my own. So if I have something that's found many times and has a zoom, but no download, what should I do? I can understand looking at history and what's searched or what's more viewed, but how does that help me sell more?

As far as the image rank, or I should say pseudo rank, there are people on the Alamy forum that track and watch and say that there hasn't been a re-rank in a long, long time. Unlike other places,if things are ranked anymore they are by Pseudo.

I too wonder why some really work the AoA data.  It's too much work for me.

My secret tweak is as follows:  I have two pseudos.  (In prior days psuedo rank was really important for search placement.  Still seems to be although it doesn't vary much anymore.)
My top pseudo consists of sold or zoomed images.  It has a high resulting rank.  So when an image is sold or is zoomed, I change the pseudo for the image.
This is also a fair bit of work, but keeping it up daily when I check sales, it's easy.

Interesting idea and that makes sense. Is there any way you can see that that effort actually does anything? I mean if I put only my sold and popular images in the one pseudo, isn't that kind of like saying, after the fact, that something was a success? How do I predict the future?  :)

Maybe what I do, doesn't lend itself to repeat sales from multiple buyers?

« Reply #4 on: April 18, 2022, 07:59 »
0

Interesting idea and that makes sense. Is there any way you can see that that effort actually does anything? I mean if I put only my sold and popular images in the one pseudo, isn't that kind of like saying, after the fact, that something was a success? How do I predict the future?  :)

Maybe what I do, doesn't lend itself to repeat sales from multiple buyers?

A new pseudo will start in a mid-rank position.  When pseudo rank moved around, rather than picking my "best" shots for my "A grade" pseudo, I let the customer decide, using sales or zooms.  Now, with stable pseudos, I'm not so sure what impact this has.  But as you can imagine, it turns out that a popular image with one client is likely to be more popular with the next one.   Also, it does make searching for usage for payback schemes easier since most sales are under my top pseudo name.

Uncle Pete

  • Great Place by a Great Lake - My Home Port
« Reply #5 on: April 18, 2022, 11:39 »
0

Interesting idea and that makes sense. Is there any way you can see that that effort actually does anything? I mean if I put only my sold and popular images in the one pseudo, isn't that kind of like saying, after the fact, that something was a success? How do I predict the future?  :)

Maybe what I do, doesn't lend itself to repeat sales from multiple buyers?

A new pseudo will start in a mid-rank position.  When pseudo rank moved around, rather than picking my "best" shots for my "A grade" pseudo, I let the customer decide, using sales or zooms.  Now, with stable pseudos, I'm not so sure what impact this has.  But as you can imagine, it turns out that a popular image with one client is likely to be more popular with the next one.   Also, it does make searching for usage for payback schemes easier since most sales are under my top pseudo name.

Maybe I should make a most popular Pseudo except my best selling images all have one download.  ;) Kind of specialized collection and limited use, where someone wants something and gets it and no one else wants the same.

I'm just wondering about the general usefulness of the measure for ALL and I think everyone who answered looks at the same as I do. ALL isn't of use, but our own measures can be.

Yes to rank and re-rank. I have looked and made notes and made more notes, and also played with keyword order and some other ways to influence my images vs my images. I don't see much of great interest or taking a lot of time. Basically upload, use the best description I can come up with, best keywords that apply, and let them fly.

I can say that after some recent testing, and what someone else advised me, Good accurate Descriptions (aka captions or titles) are important on Alamy. Possibly more important than keywords in some cases. Use those 150 words carefully?

https://www.alamy.com/myupload/help/AIM-InstructionManual.pdf#page=16

2. The Caption should be a concise and accurate description of your
image. A good caption not only helps with search relevancy on Alamy,
but also has wider benefits for search engine optimisation meaning
visitors searching for relevant subjects on search engines are also more
likely to find your image


It is very simple.  If you look whole Alamy you will get overall trends, but there will be lots of data.  I generally look at my own more.

Thank you and thanks for leading me to verify the Captions importance.

zeljkok

  • Non Linear Existence
« Reply #6 on: April 18, 2022, 13:49 »
+2
Once again, to sum up short what I posted earlier:   Main benefit of Alamy measures is ability to come up with optimal set of Metadata -  Title (Caption),  10 super, Rest.   This in turn generates better CTR, which props Alamy Rank.   

Importance of proper keywording can not be overstated, and by examining Measures you get insight into what real people are looking for, as opposed to AI generated "suggestions" on other sites which is often totally off.   I sometime update Metadata on AS too, based on keyword(s) I learned about on Alamy measures. 

Uncle Pete

  • Great Place by a Great Lake - My Home Port
« Reply #7 on: April 18, 2022, 16:41 »
0
Once again, to sum up short what I posted earlier:   Main benefit of Alamy measures is ability to come up with optimal set of Metadata -  Title (Caption),  10 super, Rest.   This in turn generates better CTR, which props Alamy Rank.   

Importance of proper keywording can not be overstated, and by examining Measures you get insight into what real people are looking for, as opposed to AI generated "suggestions" on other sites which is often totally off.   I sometime update Metadata on AS too, based on keyword(s) I learned about on Alamy measures.

Yes to that all except for Alamy rank, which may be imaginary. My images have not moved up or down in any search in over a year.

""suggestions" on other sites which is often totally off" and blind leading the blind to add more words that came from the blind. I can learn something from SS words used to find this image, for my images that sell, and Alamy words used to find my images.

I suppose if I wanted to be a spam master, I could just include most searched words from ALL, into some of my Alamy images. I'd get views and no sales, but darn, that's what people do on other sites too. More word, more words, get views, even if the image isn't a match? How did that become part of the Microstock game? I won't do that, but we all know, you must have 50 words on every image to make more sales!  ::)


zeljkok

  • Non Linear Existence
« Reply #8 on: April 18, 2022, 23:29 »
0

Yes to that all except for Alamy rank, which may be imaginary.

Alamy Rank is very real.  Click Through Rate (CTR) is just one of factors in play, they won't disclose others but sale frequency/amount likely comes into play as well.  Reason I love CTR False Positives concept is because it effectively fights keyword spamming.  Put the word donkey into image of polar bear, fine.  But when it gets returned in searches nobody will zoom it, thus lowering your Alamy Rank & impacting future sales.

Uncle Pete

  • Great Place by a Great Lake - My Home Port
« Reply #9 on: April 19, 2022, 11:42 »
0

Yes to that all except for Alamy rank, which may be imaginary.

Alamy Rank is very real.  Click Through Rate (CTR) is just one of factors in play, they won't disclose others but sale frequency/amount likely comes into play as well.  Reason I love CTR False Positives concept is because it effectively fights keyword spamming.  Put the word donkey into image of polar bear, fine.  But when it gets returned in searches nobody will zoom it, thus lowering your Alamy Rank & impacting future sales.

Shouldn't my images move up or down then? They are the same after years.

By the way, in case someone is new and wonders, I asked Alamy about rank and the answer was, "everyone starts in the middle". I don't know how that translates into, what is the middle. In my imagination, if there are 99 photos of something, and they are ranked high to low, my new image will be placed at #500.

But lets be more realistic, middle isn't center point, it's middle rank. The 999 images could be 666 ranked above average and 333 below, so my new image would be placed at 334. I think that's more logical than just plop and I'm at the middle of all images.

Ah but Alamy doesn't use individual IMAGE rank, they use artist rank? So my overall rank for my images for that pseudo, is where my next image gets slotted in. AS an artist I could be middle and still middle. Another answer from Alamy was "you don't have enough images and sales to be ranked." I'm middle.  8)

Remember that code word test that Alamy artists used? BHZ The test was, a commercial photograph, RM, and see what page you landed on, compared to all others who included that word, in the box one keywords. Well no boxes, so I'd suppose a standard keyword, nothing Supertag and that game could still be played. BHZ Stock Photos and Images (3,076)

Hey and there I am, page 14 photo 81. Still on page 14 after all these years.

Quote
Alamy Admin
 Posted June 4, 2014

There is no way to 'game' the system.

You get a good rank by submitting pictures that generate sales and zooms. Good keywording is important because without that, your images don't get seen.

So, the only thing you can do is submit your best work and keyword it well.

And BHZ rerank:  https://discussion.alamy.com/topic/6410-rerank/

Is this good or bad?

Top performing pseudonym: PETE KLINGER(CTR: 0.86)
Last month average CTR on Alamy: 0.58

Better than average but so what? One crummy $12 sale.
« Last Edit: April 20, 2022, 08:40 by Uncle Pete »

« Reply #10 on: April 25, 2022, 00:35 »
0
Haven't really bothered with Alamy Measures or CTR. Maybe more attention to them would help or then not, can't really say. How does the system know who clicks your image - is it a buyer or somebody looking for suitable UK keywords, or just someone on Google looking? Clicks are not equal to sales.
Only tried to caption and keyword as accurately as possible, and upload relevant images.

Uncle Pete

  • Great Place by a Great Lake - My Home Port
« Reply #11 on: April 25, 2022, 09:00 »
0
Haven't really bothered with Alamy Measures or CTR. Maybe more attention to them would help or then not, can't really say. How does the system know who clicks your image - is it a buyer or somebody looking for suitable UK keywords, or just someone on Google looking? Clicks are not equal to sales.
Only tried to caption and keyword as accurately as possible, and upload relevant images.

Registered buyers, not just random people coming to look.

I think there's good to be seen from views, no money, but I can see what people are looking at, and maybe guess if I wanted to make more of that? Zooms could tell us, that someone looked close and didn't buy the image. Does that mean it didn't meet their needs or something else? We don't know. But that could be a lead for things to shoot. Again, only within my own. The whole all of Alamy is a waste of time, and most of those you can see were a project that someone needed images, specific images, and they probably won't need any more ever again.

I think Alamy rank is over credited with meaning something to artists.

Remember Alamy diversity, where roughly speaking, you can't have a second image showing for the same search, until all the other artists have had one image, shown fro that search. Sometimes that means my image is 8th sometimes 14th and for some specific words and locations, I'm first. Alamy diversity in the search means more for us than rank in most cases.

I suppose if it was a very popular subject and a terrible two word search, we might end up on page 2 or 3, but a good thoughtful search with at least three good words, being a good match, we'd be seen close to the front.

Isn't the whole rank concern and all of that, that people worry about so much, about being seen on the first page or second page? Go look... we are, with good titles and good accurate keywords, and the diversity algorithm.

Uncle Pete

  • Great Place by a Great Lake - My Home Port
« Reply #12 on: April 25, 2022, 09:02 »
0
Into the past:

Hello everyone!

Apologies for the confusion / mystery surrounding 'AlamyRank' - truth is, it's a very complex algorithm and not a simple forumla - also, because we need to make tweaks and improvements to the search functionality as time goes on, there is a certain amount of fluidity to the precise workings of it all. CTR, Sales history, Keyword placement (essential, main, comp) and keyword proximity (where words appear within the search box) also come into it as one big mix that gives your images a 'score' which is then fed into the search results. This all happens in fractions of a second to return images in what is probably the fastest stock image search engine in the world.

Put simply though, the only tactic you can employ is to submit your best work and keyword it thoroughly with a high degree of relevancy.

Thinking specifically about CTR then. CTR (Click Through Rate) does have an impact on your ranking (as part of the big mix mentioned above), and we give you the average figure across all of Alamy so you can see how your images are performing compared to your competition.  If your images are seen many times in search results but never clicked on, this will have a negative effect on your rank and vice-versa. The search engine will also mix some results around to give everyone a fair chance, particularly if you are just starting out and have the median rank. Sales and the sales value have positive effect on your rank, so the more sales you make and the higher the value those sales are contribute significantly.

We are also doing lots of work behind the scenes on the search results:

http://www.alamy.com/Blog/contributor/archive/2012/10/19/5164.aspx

Further info here:

http://www.alamy.com/Blog/contributor/archive/2012/10/29/5241.aspx

...and this is all yielding positive results for our customers:

http://www.alamy.com/Blog/contributor/archive/2012/12/20/5288.aspx

So hopefully this clarifies things a little for you. In  a nutshell - CTR is important, but so are sales, keyword relevancy, Keyword proximity and all the workings the content team are doing behind the scenes on developing the search results.

Oh, and a couple more points for you that will hopefully be useful!

- "I had a sale but no zoom"

We can only match up a zoom to a sale if the customer completes this in one go, eg, search, zoom, buy. Quite often the customer will search and then come back later to buy, so we can't match these two up. Also, sometimes the picture researcher who finds the images, then passes them over to another account for purchasing, again, we can't match these two up as related.

- Play the "BHZ" game

Our advice on this? Don't bother. It only tells you how well you are doing compared to everyone else playing the BHZ game and it is very inconsistent because some people have multiple images in there, some have BHZ in the caption only, some in the essential keyword field, some only in the comprehensive field etc etc. Combine that with the 'big mix' (above) of elements that determine the ranking, you can quickly see that your BHZ placement is going to tell you virtually nothing. The better way to see how you are performing is to keep an eye on some real searches and see where you are coming up within those results.

Finally, in terms of frequency of updating the ranks, it's not set in stone but it works out that roughly every 100 days we will update the results using the previous 300 days data.

Hope that helps!

James Allsworth
Content Executive
Alamy


This could have changed but it's from someone who knows. The links don't appear to lead anywhere except 403 errors. I didn't spend time finding a new link.

CTR is important, but so are sales, keyword relevancy, Keyword proximity

Alamy does not stem words, which means we need to do that if it's necessary and appropriate.
« Last Edit: April 25, 2022, 09:04 by Uncle Pete »

« Reply #13 on: April 25, 2022, 10:57 »
0
I just go by the zooms because hopefuly sales will follow and recently had a lot more as my alamy port aproached 1000 images.

ShadySue

  • There is a crack in everything
« Reply #14 on: April 25, 2022, 16:36 »
0
I'm not sure that the search algorithm has changed for quite some time.
Never forget that any word/s in the description can be searched for on their own or combined with any keyword/s. All of your words might be relevant, but the combined phrase may well be irrelevant, and it's not your fault.

For example, if what I wrote above was an Alamy description, it might show up in a search for your fault or irrelevant algorithm or search phrase

And if you captioned a "crowd waiting to see Emmanuel Macron after he was returned as French President", that could show up on a search for Macron, even though he isn't in the image.

zeljkok

  • Non Linear Existence
« Reply #15 on: April 25, 2022, 17:23 »
0


And if you captioned a "crowd waiting to see Emmanuel Macron after he was returned as French President", that could show up on a search for Macron, even though he isn't in the image.

Yes, and that is main reason for "false positives".  It is direct consequence of fact they index the Caption. Because it is impossible to know what should be treated as phrase since Caption is array of words. 

Different scenario is of course with tags where you can directly specify what is the phrase, and what is single indexable word.   

It is important to be aware of the above for optimal metadata specification.  I suspect most people don't fully understand this, or simply don't bother.

« Reply #16 on: April 25, 2022, 18:21 »
0
I just go by the zooms because hopefuly sales will follow and recently had a lot more as my alamy port aproached 1000 images.

I have a slightly larger sized Alamy portfolio and have been a contributor for a number of years.

It's been my experience for some time that the majority of my image sales by far drop in seemingly at random and have not been zoomed prior to sale.

What does that mean?
 

zeljkok

  • Non Linear Existence
« Reply #17 on: April 26, 2022, 01:27 »
0
I just go by the zooms because hopefuly sales will follow and recently had a lot more as my alamy port aproached 1000 images.

It's been my experience for some time that the majority of my image sales by far drop in seemingly at random and have not been zoomed prior to sale.

What does that mean?
 

Most contributors share that experience;  Alamy is highly unpredictable.

Re most sales not being zoomed:  Zooms are captured only from subset of customers,  how many I am not sure but it is way below 50%.   

Uncle Pete

  • Great Place by a Great Lake - My Home Port
« Reply #18 on: April 26, 2022, 11:03 »
0
I just go by the zooms because hopefuly sales will follow and recently had a lot more as my alamy port aproached 1000 images.

It's been my experience for some time that the majority of my image sales by far drop in seemingly at random and have not been zoomed prior to sale.

What does that mean?
 

Most contributors share that experience;  Alamy is highly unpredictable.

Re most sales not being zoomed:  Zooms are captured only from subset of customers,  how many I am not sure but it is way below 50%.

Also as explained by Alamy, in the quote that I just posted yesterday? 

"We can only match up a zoom to a sale if the customer completes this in one go, eg, search, zoom, buy. Quite often the customer will search and then come back later to buy, so we can't match these two up. Also, sometimes the picture researcher who finds the images, then passes them over to another account for purchasing, again, we can't match these two up as related."

Zooms are good information for what's people are really looking at.

I'm not sure that the search algorithm has changed for quite some time.
Never forget that any word/s in the description can be searched for on their own or combined with any keyword/s. All of your words might be relevant, but the combined phrase may well be irrelevant, and it's not your fault.

For example, if what I wrote above was an Alamy description, it might show up in a search for your fault or irrelevant algorithm or search phrase

And if you captioned a "crowd waiting to see Emmanuel Macron after he was returned as French President", that could show up on a search for Macron, even though he isn't in the image.

And the choice is not have those description words searched? So why have descriptions at all, anywhere?

There's also something else as long as this is turning into how does the Alamy search work? Word proximity which is also kind of obscure. But words in the right order, closer together are stronger than the same words, not in the searched order and not close to each other. I don't know if anyplace else does that?

I'm happy that they dropped the three box system!

I think I'm learning better to look at views and only my own data.

zeljkok

  • Non Linear Existence
« Reply #19 on: April 26, 2022, 14:10 »
0
So there is interesting "False Positive" I am seeing today.  This is the clicked image:


Customer Search Phrase was  "Vancouver Washington", which is the City in US Washington State.   This has NOTHING to do with the image which is Waterfront view of Strait of Juan De Fuca on Saxe Point public park near Victoria BC on Vancouver Island, Canada.   I had "Vancouver Island" as Tag, and "Distant Olympic Peninsula Washington USA" part of Caption. 

So their search engine took "Vancouver" word from "Vancouver Island" phrase, joined it with word "Washington" from Caption and made a match!  To make things even more interesting,  it generated a click ?!

I probably should not have "Washington" in Caption, and I am going to remove it now, but still the click ?!  We are probably sometimes overthinking the whole thing

« Reply #20 on: April 26, 2022, 14:59 »
0
When I see searches for something I have in my portfolio, it's usually just a prompt for me to finish processing images from a shoot & I'll add some new or additional images in case the person searching is still looking.

I can only recall one real success. Right before a 2016 photo trip to Cape Cod, someone zoomed a 10-year-old image of a specific building. I was staying with a friend in the town where I'd taken the original image, so I shot several new pictures when I arrived and uploaded them right away. Sold one of the new ones for $250 on Alamy less than a month later. It was really just pure luck but that single sale pretty much paid for my 5-day trip: only gas & food since I was staying with a friend.

What paid off more was that, if I hadn't had that zoom, I would have concentrated on beach, boat and lighthouse images, evergreen sellers for me, not new shots of a historic building. While photographing that building, I also shot a lot of the cute little town, photos I've licensed to calendar companies, as stock, and sold as fine art. It got me thinking differently. A few of the images have been shown in galleries here in NY and in New England too.

Anything that gets your creative juices flowing is worthwhile.

I've gotten some very discouraging low prices on Alamy of late, but I see from the Alamy forum that there are still some $$$ licenses to be had even now, and high $$ are still frequent enough to make it worthwhile (I got a couple $70+ last month).


 

Related Topics

  Subject / Started by Replies Last post
11 Replies
7134 Views
Last post December 19, 2007, 22:28
by sharpshot
26 Replies
15592 Views
Last post October 28, 2008, 16:25
by RacePhoto
3 Replies
3139 Views
Last post February 19, 2012, 09:56
by Eireann
1 Replies
2677 Views
Last post June 16, 2013, 19:55
by klsbear
22 Replies
8364 Views
Last post November 29, 2018, 12:33
by Uncle Pete

Sponsors

Mega Bundle of 5,900+ Professional Lightroom Presets

Microstock Poll Results

Sponsors

3100 Posing Cards Bundle