When was the last re-rank?

[M.Chapman]

Does anyone know when the last re-rank was?

I haven't noticed any evidence of a re-rank since the new AIM and data structures were introduced (about 2 years ago?).

I’ve asked James Allsworth at Alamy about this but haven't received a clear answer. If re-ranks aren’t happening then I worry that this undermines one of the basic mechanisms I've long believed can be used to improve the position of my images in search results. i.e. maximise CTR% by tight and accurate keywording, thereby reducing the chance of getting unwanted views and increasing zooms as a proportion. I also worry that if re-ranks (in which CTR% is one of the ingredients) aren’t being carried out, then there’s no penalty for keyword spamming and customers will see more irrelevant images, and there's less potential of reward to contributors who work hard to improve their CTR%. 

[Allan Bell]

On 10/05/2020 at 09:30, M.Chapman said:

Does anyone know when the last re-rank was?

I haven't noticed any evidence of a re-rank since the new AIM and data structures were introduced (about 2 years ago?).

I’ve asked James Allsworth at Alamy about this but haven't received a clear answer. If re-ranks aren’t happening then I worry that this undermines one of the basic mechanisms I've long believed can be used to improve the position of my images in search results. i.e. maximise CTR% by tight and accurate keywording, thereby reducing the chance of getting unwanted views and increasing zooms as a proportion. I also worry that if re-ranks (in which CTR% is one of the ingredients) aren’t being carried out, then there’s no penalty for keyword spamming and customers will see more irrelevant images, and there's less potential of reward to contributors who work hard to improve their CTR%. 

 

I think this was asked in another thread somewhere and the consensus seemed to be that Alamy are not re-ranking like they used to do long ago.

 

It could have even been mentioned by alamy at some point if my memory serves me right.

 

Allan

 

 

[M.Chapman]

On 11/05/2020 at 10:52, Allan Bell said:

 

I think this was asked in another thread somewhere and the consensus seemed to be that Alamy are not re-ranking like they used to do long ago.

 

Allan

 

Which, if true, means that the green discoverability bar could be useful after all, as there will be NO penalty for keyword spamming. Extra keywords will generate extra views and may generate a slightly increased chance of selling. The click through rate (CTR%) will fall but, if re-ranks aren't happening, the contributor's rank will never suffer so their images won't get demoted.

 

Similarly - Contributors who currently have a low rank have no chance to improve it. If their current rank is below average, their best tactic maybe to create a new pseudo (which will have average rank) and move their images over to it. It's simple to try that test. My main pseudo has above average rank (last re-rank I can see was around Jan 2018). If I create a new pseudo and move an image to it, then its position tends to drop in search results (assuming those searches create lots of hits). If I move it back, then its position is restored. NB. Alamy can change their algorithms at any time...

 

Mark

Edited May 14, 2020 by M.Chapman

[Allan Bell]

2 hours ago, M.Chapman said:

 

Which, if true, means that the green discoverability bar could be useful after all, as there will be NO penalty for keyword spamming. Extra keywords will generate extra views and may generate a slightly increased chance of selling. The click through rate (CTR%) will fall but, if re-ranks aren't happening, the contributor's rank will never suffer so their images won't get demoted.

 

Similarly - Contributors who currently have a low rank have no chance to improve it. If their current rank is below average, their best tactic maybe to create a new pseudo (which will have average rank) and move their images over to it. It's simple to try that test. My main pseudo has above average rank (last re-rank I can see was around Jan 2018). If I create a new pseudo and move an image to it, then its position tends to drop in search results (assuming those searches create lots of hits). If I move it back, then its position is restored. NB. Alamy can change their algorithms at any time...

 

Mark

 

Thanks Mark. I had not really thought that one through. Interesting thoughts on your part.

 

Allan

 

[gvallee]

According to the BHZ game, my ranking has been exactly the same since November 2017. I haven't checked with real searches though.

[M.Chapman]

18 minutes ago, gvallee said:

According to the BHZ game, my ranking has been exactly the same since November 2017. I haven't checked with real searches though.

 

You're right. Looking at my last change according to the BHZ test it was 25/11/2017. Other measures (including rolling average of views/week expressed as a percentage of my portfolio) also indicate that nothing significant has changed for me since around Jan 2018 (slight lag due to the rolling average).

 

Mark 

Edited May 14, 2020 by M.Chapman

[NYCat]

They always said they wanted to do rankings for individual images and I think that is happening to a certain extent with mine. I am fortunate to have a good ranking overall and I guess I had to have had that when they stopped the big re-ranking.

 

Paulette

[Betty LaRue]

Lots of uncertainty. The thing is, other than occasional temporary spikes in my CTR, it has always been below average. Yet when I used to do searches (I don’t bother much anymore) my images were first page often, in first few pages for others.

My CTR has actually improved lately.

What it all means, to figure it out with any certainty, is above my pay grade. 😎

[geogphotos]

1 hour ago, Betty LaRue said:

Lots of uncertainty. The thing is, other than occasional temporary spikes in my CTR, it has always been below average. Yet when I used to do searches (I don’t bother much anymore) my images were first page often, in first few pages for others.

My CTR has actually improved lately.

What it all means, to figure it out with any certainty, is above my pay grade. 😎

 

 

I never look at CTR. Not suggesting that is sensible just a fact.

 

And as you say Betty when things become so uncertain you wonder what is the purpose of all these measures. Actually we have measures which seem to conflict with each other - make your images more 'discoverable' and get a green reward, only to see your CTR go down.

 

I'm not surprised if it is true that Alamy have stopped re-ranking, and even less surprised that they have kept quiet about whatever they have done.

[geogphotos]

On 10/05/2020 at 09:30, M.Chapman said:

Does anyone know when the last re-rank was?

I haven't noticed any evidence of a re-rank since the new AIM and data structures were introduced (about 2 years ago?).

I’ve asked James Allsworth at Alamy about this but haven't received a clear answer. If re-ranks aren’t happening then I worry that this undermines one of the basic mechanisms I've long believed can be used to improve the position of my images in search results. i.e. maximise CTR% by tight and accurate keywording, thereby reducing the chance of getting unwanted views and increasing zooms as a proportion. I also worry that if re-ranks (in which CTR% is one of the ingredients) aren’t being carried out, then there’s no penalty for keyword spamming and customers will see more irrelevant images, and there's less potential of reward to contributors who work hard to improve their CTR%. 

 

 

I'm not sure that it is in Alamy's best interests ( or indeed the client who is searching) to give too much 'power' to individual contributors over search position. With 200 million images the customer has to learn how to search and I'm sure that they can do that. Alamy gets the same commission whoever gets the sale, and  you could argue that the large libraries who do not have time to 'play the system' are a short-cut simple way to offer what the punters want. By contrast why give 'power' to thousands of contributors ( a totally mixed ability bunch to start with) to shove their images up to the top? Actually, even more than that it is giving disproportionate influence to the relatively small number who have the time and knowledge to play the game to their own advantage.

 

Maybe we need a new version of BHZ ( Tony Collins your fame lives on)? One that we could play between ourselves on the forum to see how our ranks compare and to act as a test-bed for tweaks? It would seem obvious to call it the ZHB game 😜

Edited May 14, 2020 by geogphotos

[M.Chapman]

8 hours ago, NYCat said:

They always said they wanted to do rankings for individual images and I think that is happening to a certain extent with mine.

 

I can see that is happening with my images to some extent. If an image is zoomed, its position is improved in search results if the same search term is used again. So zooms are good for 3 reasons. Not only do they indicate extra interest in an image (which might lead to a sale), the position of that image is improved in subsequent searches and they drive up CTR% (which is potentially used as one of the factors when calculating a contributor's rank). I've not noticed the same effect with images that have sold, but that's no so easy to analyse. 

 

But overall, when my BHZ test position has fallen I see a reduction in views (indicating my images are, on average, appearing lower down in search results), and when my BHZ test position has risen, I see a rise in views. But the position in which any individual image might appear in a search depends on many factors including my rank versus the rank of the other contributors whose images also meet the search criteria; and how each of us has used tags, captions and supertags; and on previous zoom history; and potentially other stuff too, like date taken. It's very complex, but I see rank as very important because it potentially affects the position in which all my images appear in searches. Whereas zoom history (on an individual image) only affects the images which have been zoomed (around 10% of my portfolio). Sales history of an individual image would effect even fewer images in my case. 😞

 

That's why I'm concerned that I can't see any evidence of a re-rank in my data for over 2 years now. 

 

NB. Usual caveat applies. Alamy can change their search algorithm at any time.

 

Mark

Edited May 14, 2020 by M.Chapman

[M.Chapman]

2 hours ago, geogphotos said:

Actually, even more than that it is giving disproportionate influence to the relatively small number who have the time and knowledge to play the game to their own advantage.

 

Is accurate, concise keywording a "game"?

Shouldn't contributors that get this right, and maximise zooms per view (which is presumably good for Alamy and their customers) be additionally incentivised.... 

 

Mark

Edited May 14, 2020 by M.Chapman

[geogphotos]

32 minutes ago, M.Chapman said:

 

Is accurate, concise keywording a "game"?

Shouldn't contributors that get this right, and maximise zooms per view (which is presumably good for Alamy and their customers) be additionally incentivised.... 

 

Mark

 

You would have to ask Alamy.

 

But my guess is that, after all said and done, they prefer not to allow the 'average' contributor to have too much direct control over search engine position.

 

I'm sure that the answer would be that 'accurate concise keywording' is expected of everybody and has its own reward.

 

 

Edited May 14, 2020 by geogphotos

[Johnnie5]

I have noticed that photos of mine that have previously sold place higher in searches versus photos that have never sold.  Maybe some of you could experiment and see if it is the same for yourselves.

[M.Chapman]

If the image that sold was also zoomed, then I see it appear higher relative to my other images of the same subject (caption and tag differences aside). But if it’s just a sale, I haven’t noticed any improvement in position, indeed some of my sold images appear last. But it’s not something I’ve looked at closely, and Alamy can, and do, change the weightings the search algorithm applies to each factor from time to time.

 

Mark

Edited May 15, 2020 by M.Chapman

[Jansos]

Lots of talk on here about CTR, my CTR hovers somewhere between 0.45 - 0.65. Average CTR on Alamy often seems to be about 0.59 ish. As a matter of interest what CTR do regular contributors experience? Obviously above average is good but what type of CTR makes a difference? Any ideas?
BW, John

Edited May 29, 2020 by Jansos
typo

[Steve F]

41 minutes ago, Jansos said:

Lots of talk on here about CTR, my CTR hovers somewhere between 0.45 - 0.65. Average CTR on Alamy often seems to be about 0.59 ish. As a matter of interest what CTR do regular contributors experience? Obviously above average is good but what type of CTR makes a difference? Any ideas?
BW, John

When I started off I was well above average on CTR, but now I'm on a 0.54 rolling monthly average (depressingly below Alamy's 0.59 average). I probably need to go back and look at some of my earlier captions and keywords at some point.

 

It's probably very difficult to compare contributors because you can't look at the stats for a start off. Plus it's all a function of how big your collection is, how often you submit new material, how often you sell, CTR, subject matter, location, image quality etc.....

Steve

Edited May 29, 2020 by Steve F

[Allan Bell]

My average CTR for the last year is 0.48 highest being 1.10 lowest 0.00

 

Allan

 

 

 

[geogphotos]

If there are no Re-Ranks  then why worry about CTR?

 

The larger the entire collection gets the more meaningless this Click Through Ratio becomes.

 

You could well argue that to more attract sales you should offer more image choice rather than less and accept a low CTR.

Edited May 29, 2020 by geogphotos

[M.Chapman]

9 hours ago, Steve F said:

Plus it's all a function of how big your collection is, how often you submit new material, how often you sell, CTR, subject matter, location, image quality etc.....

 

CTR is a function of CTR?? Well sort of, but only if reranks are happening....

 

Current CTR% = This month's Zooms/Views

 

Views are influence by Alamy Rank, which (if it's ever recalculated again) depends on historical CTR% (and other stuff)

Alamy Rank is believed to depend on average CTR%, sales, and probably other unknown stuff as recorded over some earlier period (6 months?) 

Alamy Rank influences views. The higher a contributor's rank, the higher their images appear in search results, so the more views they are likely to receive.

But views are also affected by the following - Subject matter (are customer's searching for subjects in a contributor's portfolio, and how much competition is there on Alamy), contributor's portfolio size (more images = more views) and keywording (more keywords = more views).

Zooms indicate customers are interested in the images they have seen.

If reranks are happening it's believed to be good to have a high CTR% (Alamy, helpfully provide a graph of CTR%). High CTR% can be achieved by striving to ensure customers are only shown images from your portfolio that they might be interested in and might zoom.

To optimise CTR%

- Use concise accurate keywords (to avoid irrelevant views)

- Avoid uploading duplicates (the images may be good, but it's likely only one will be zoomed)

- Only upload best images

- Pick subject matter with minimal competition 

 

That's the theory anyway. But, unless Alamy re-ranks are reinstated, zooms/view (=CTR%) is a bit academic. Zooms are still good though, as they indicate customers are interested in a contributor's images, and zoomed images may be promoted in future searches. It's also worth avoiding uploading duplicates and only uploading the best images because it's not easy to control the order in which images will appear in search results. (Upload too many of the same subject and the best image might appear last and never be seen). 

 

Mark

[Jansos]

Thanks for all your replies - most helpful. :-)

[Steve F]

1 hour ago, M.Chapman said:

 

CTR is a function of CTR?? Well sort of, but only if reranks are happening....

 

Ha ha, you got me, I should have said - how successful you are is likely a function of how big your collection is, how often you submit new material, how often you sell, CTR, subject matter, location, image quality etc..... CTR is just one measure we know about.

Edited May 29, 2020 by Steve F

[Allan Bell]

On 29/05/2020 at 15:26, Allan Bell said:

My average CTR for the last year is 0.48 highest being 1.10 lowest 0.00

 

Allan

 

 

 

 

This is new for me. The CTR graph in Alamy Measures in my dashboard has gone up to 13.33 today.

 

Yes I have read it right 13.33??

 

Allan