BHZ changed for the first time in years

[Steve Dorey]

Hi All

I don't know how many of you still use bhz to measure your position in searches.

I'm a bit of a bean-counter so I have measured my different pseudonyms on the 1st of every month since 2007.

(It's a bit sad really)!

Anyway, today I have noticed the first significant change in the position of some of my pseudonyms since 2017.

(Alamy used to update them every 100 days or so).

Has anyone else noticed a change?

 

[geogphotos]

31 minutes ago, M.Chapman said:

Obviously Alamy can do whatever they want. But it's hardly a sound basis for long term collaboration if a contributor can steadily build a collection of images that are selling well, only to find Alamy then change their rank so that their images are less likely to be seen.... Alamy's lack of transparency on the contributor ranking system has always been a concern for me as it means that they could suddenly change my ranking and hence income, without any consultation or explanation.

 

Mark

 

 

 

Your concerns may be very valid. Who knows? 

 

Alamy does not offer transparency. It offers a platform for images to be licensed. It is a business and obviously our choice whether to take part.

 

At one stage of the business Alamy portrayed itself as the 'photographer's friend' and many of us were enticed by that. Those days are long over. 

 

Yes, of course, Alamy will change the ranking system as they see fit without consultation or explanation. That was never promised. 

Edited May 12, 2022 by geogphotos

[Reimar]

I thought I'd check another image in my top pseudo with a BHZ supertag to see where it lands.  As I recall, they used to be quite close to each other.  This one landed on page 3 instead of 7 when including RF images since I forgot to make sure I picked a RM image.  More testing to come....

[Reimar]

Just to finish that thought:  a RM image previously sold and sold again yesterday placed on page 3 rather than page 6 for an image that hasn't sold for years.  So in addition to pseudo rank, individual image CTR/sold status must be having an effect.  Last time I looked many years ago individual images of a pseudo clustered around the same page.

[wiskerke]

23 hours ago, Reimar said:

I thought I'd check another image in my top pseudo with a BHZ supertag to see where it lands.  As I recall, they used to be quite close to each other.  This one landed on page 3 instead of 7 when including RF images since I forgot to make sure I picked a RM image.  More testing to come....

 

2 hours ago, Reimar said:

Just to finish that thought:  a RM image previously sold and sold again yesterday placed on page 3 rather than page 6 for an image that hasn't sold for years.  So in addition to pseudo rank, individual image CTR/sold status must be having an effect.  Last time I looked many years ago individual images of a pseudo clustered around the same page.

 

So BHZ still has it's value: as a testing ground to answer questions such as this.

Because the algo has gotten much more complicated, testing has become more complicated also.

Try the same method for the keywords that sold your image.

 

wim

[geogphotos]

1 hour ago, wiskerke said:

 

 

So BHZ still has it's value: as a testing ground to answer questions such as this.

Because the algo has gotten much more complicated, testing has become more complicated also.

Try the same method for the keywords that sold your image.

 

wim

 

 

The same sort of testing of image rank - what difference does it make if I add a Supertag, what happens if I remove the tag completely and leave it in caption only -  can surely be done more effectively using real life searches against other images that won't have made the same changes overnight. 

 

I must admit that I don't get around to adding Supertags and just hope for the best. 

Edited May 14, 2022 by geogphotos

[John Mitchell]

4 hours ago, Reimar said:

Just to finish that thought:  a RM image previously sold and sold again yesterday placed on page 3 rather than page 6 for an image that hasn't sold for years.  So in addition to pseudo rank, individual image CTR/sold status must be having an effect.  Last time I looked many years ago individual images of a pseudo clustered around the same page.

 

I haven't done any testing, but I think you're correct. In my experience, image sold status has a big effect when it comes to placement of individual images -- i.e. if an image has been licensed, it is much more likely to appear on the first page of relevant search results.

[Gorilla Dave]

I got interested in the BHZ game again…

I did some tests and thought I’d share the results in case anyone’s interested.
This is for an image that has never sold but it’s in a pseudo that has fared at least average.

BHZ placed in…                image found on page…

 

Normal tag only                             p28  (of 31)
Caption only                                   p26
Supertag only                                  p11

 

Caption and normal tag                            p24
Supertag and normal tag                           p 7
Caption and supertag                                 p 2

 

Caption, supertag and normal tag            p 1

 

 

Hope some folks might find this useful.

 

I think it shows it's well worth putting some thought into your choice of supertags;

then include all your supertage as normal tags also.

Then make full use of the caption, and include a phrase or phrases in the caption field that make use of all the supertags.

 

Also - double-check all your spellig !

 

GD

 

[Harry Harrison]

1 hour ago, Gorilla Dave said:

Caption, supertag and normal tag            p 1

 

Very interesting, I certainly haven't kept up with all the nuances of the BHZ game but I don't remember reading this before. I may be misunderstanding you (and of course I can try it myself) but it seems quite strange (and counter-intuitive) that a keyword/tag needs to be entered twice, once as a supertag and once on its own. However, as anyone who has downloaded the data csv will know, the tags and supertags are separate data fields ('Tags' & 'Super Tags' respectively), in other words once you make a keyword a supertag it disappears from the 'Tags' field and moves to the 'Super Tags' field. It seems from your investigations that it would have been better for everyone's ranking if it was copied rather than moved, though of course that would have just levelled the playing field. It seems on the face of it like a strange anomaly of the current ranking system.

Edited May 19, 2022 by Harry Harrison

[gvallee]

5 minutes ago, Harry Harrison said:

Very interesting, I certainly haven't kept up with all the nuances of the BHZ game but I don't remember reading this before. I may be misunderstanding you (and of course I can try it myself) but it seems quite strange (and counter-intuitive) that a keyword/tag needs to be entered twice, once as a supertag and once on its own. However, as anyone who has downloaded the data csv will know, the tags and supertags are separate data fields ('Tags' & 'Super Tags' respectively), in other words once you make a keyword a supertag it disappears from the 'Tags' field and moves to the 'Super Tags' field. It seems from your investigations that it would have been better for everyone's ranking if it was copied rather than moved, though of course that would have just levelled the playing field. It seems on the face of it like a strange anomaly of the current ranking system.

 

It is not new. It was the case prior to AIM. It was a little trick not widely known.

[geogphotos]

Long, long ago, in a land before Supertags, before even BHZ was a twinkle in the eye of its creator, the Alamy search engine was a much more basic beast.

 

As an experiment I succeeded in moving one of my images to the number one spot, on page one, for 'London'. 

 

How did I do that? I added the word London more than anybody else to all the fields that were available.  I suppose the logic of the search engine meant that must be the best match. 

 

Of course, the search engine is much more sophisticated now but it still must run on computer logic - so I can quite believe that having the same keyword in tag/supertag/caption would give extra algo juice. 

Edited May 19, 2022 by geogphotos

[Harry Harrison]

16 minutes ago, gvallee said:

It is not new. It was the case prior to AIM. It was a little trick not widely known.

Thanks, well at least I've only got 2740 images to think about, and then they might just reconfigure the algorithm anyway....

 

It does seem daft.

[spacecadet]

20 minutes ago, geogphotos said:

extra algo juice. 

Mmm algo juice

[Gorilla Dave]

…it does seem counter intuitive that there should be any need or benefit from including a keyword as a normal tag AND as a supertag.

However, one benefit of doubling-up on normal tags and supertags is that it does help boost your discoverability bar into the green, and we all know how frustrating that game can be !!!!!!!!!

GD

[gvallee]

6 minutes ago, Gorilla Dave said:

…it does seem counter intuitive that there should be any need or benefit from including a keyword as a normal tag AND as a supertag.

However, one benefit of doubling-up on normal tags and supertags is that it does help boost your discoverability bar into the green, and we all know how frustrating that game can be !!!!!!!!!

GD

 

Prior to AIM, Alamy had threatened some kind of punishment for doing so. Did it ever happen? Hard to say. Duplicates did get transferred to AIM.

The colour of the discoverability bar has no effect whatsoever on anything. 

[Gorilla Dave]

36 minutes ago, gvallee said:

 

 

The colour of the discoverability bar has no effect whatsoever on anything. 

 

...no, but it is a good game !

[meanderingemu]

4 hours ago, gvallee said:

 

 

The colour of the discoverability bar has no effect whatsoever on anything. 

No immediate effect.  Might still affect in long run if increases false positives

 

[Broad Norfolk]

On 04/05/2022 at 07:46, Broad Norfolk said:

 

 

My 'worst' pseudo has always been on page 1 which I don't understand. My pseudo number two has been promoted from page 2 to page 1 which leaves my main pseudo which has dropped from page 1 to page 8. 

I'm going to keep a watch on how this affects my views, zooms, and sales. Maybe it will, maybe it won't! Time will tell. It's all still a bit of a mystery to me as to how all this is calculated.

Jim 😏

 

I've just given myself an update on what's been happening with my stuff on BHZ since 4th May. My number two and three pseudos are still both on page one whilst my main pseudo has climbed back up from page eight to page three.

Views on my main pseudo dropped by over 30% as did zooms. Sales were non-existent until 18th May and CTR was dropping. There does, however, appear to be some signs of recovery over the past week with all aspects edging in an upwards direction.

Once again time will tell but there does seems to be a little room for optimism!

Jim 😏