Commission change - James West comments

[Skyscraperfan]

31 minutes ago, andremichel said:

If I was James and was going to shake things up, (which will introduce risk and upset people), I would go for something that incentivised contributors to  build sales. It should reward recent sales growth instead of past success from many years ago.

 

Newbies could start at a low commission rate but with the right commitment and ability there should be a clear path for them to travel to higher commissions. 

 

The challenge is to come up with something that works, gives some money to alamy to use for their growth plans  and is acceptable to contributors. 

If we sell more, we already get more money. That should be incentive enough. If somebody sells photos for $1000, he gets $500. If somebody sells photos for $5000, he gets $2500. That's a big incentive for selling as much as possible. We do not need an extra punishment for those who sell less than others.

[M.Chapman]

46 minutes ago, Skyscraperfan said:

If we sell more, we already get more money. That should be incentive enough. If somebody sells photos for $1000, he gets $500. If somebody sells photos for $5000, he gets $2500. That's a big incentive for selling as much as possible. We so not need an extra punishment for those who sell less than others.

 

There may already be another incentive too? Better sellers might* be given a better Alamy Rank, so their images appear higher in search results, leading to more sales?

 

*Alamy don't disclose the algorithm, so this is guesswork on my part 

 

Mark

[andremichel]

41 minutes ago, Skyscraperfan said:

If we sell more, we already get more money. That should be incentive enough.

 

Who's to say how much incentive is enough? The current system still doesn't discourage bulk submissions of mediocre images in the hope that quantity will produce more sales. 

 

41 minutes ago, Skyscraperfan said:

 If somebody sells photos for $1000, he gets $500. If somebody sells photos for $5000, he gets $2500. That's a big incentive for selling as much as possible. We so not need an extra punishment for those who sell less than others.

 

I do not automatically mean those that sell more should automatically get a higher commissions.  

 

Admittedly I haven't explored this in any detail, and any solution I spout here will upset some people, but one possibility would be for an algorithm to calculate your commission % based on your average RPI from the previous 12 months relative to the Alamy average for last year. There would be a minimum (20%?) and a maximum (80%?) and the calculation would give your value somewhere in between.  

 

We could even get more complex by only taking into account images submitted over the last 4 years. So if you can improve your submissions and the old bad ones fall out of the equation over time. If you stop submitting, after 4 years you drop back to the newbie commission rate. 

[Skyscraperfan]

Actually some of my best selling images are among those I have already uploaded more than ten years ago. It would make no sense to me to upload photos just because we should upload a certain number of photos per year.  If those photos are worse than my old photos, that would hurt my Alamy rank. The system you propose would just make me delete my old photos and upload the same photos again, just to get a better commission.

A low commission for newbies is not the right thing, if you want to attract new photographers. Especially in the combination with exclusivity. I think Alamy had a program for students that even gave them 100% commission for a while. That's how you attract new photographers.

[M.Chapman]

25 minutes ago, andremichel said:

The current system still doesn't discourage bulk submissions of mediocre images in the hope that quantity will produce more sales. 

 

I'd like to see Alamy publicise the Alamy Rank algorithm to encourage the contributor behaviour they need (i.e. maximise revenue per image in their portfolio)

 

IMHO Alamy Rank should be based on

• a modified version of CTR% (which includes downloads and lightbox additions as well as zooms)
• and revenue (or unit sales) over the previous 6 months, divided by the number of images in a contributor's portfolio

If Alamy disclosed this part of the algorithm (in detail, for example 50% of your rank is based on.... etc.), and then made it clear how Alamy Rank affects your position in search results. This would encourage contributors to; 

• actively trim their portfolios to remove any "dross" and duplicates
• tightly keyword to improve CTR
• avoid uploading duplicates

At the moment the way Alamy Rank works is kept secret (for commercial reasons) and any info is largely "hearsay" on the forum. IMHO Alamy need to be more upfront.

 

Oh, and while Alamy are at it, remove the counter-productive "discoverability" index from AIM.

 

Mark

 

[Martin Carlsson]

4 minutes ago, geogphotos said:

If we sell more, we already get more money. 

 

Untrue.

 

Incentives?

 

2006 $18,625 from approx 6,000 images

 

2018 Now 60,000 images and circa $14k - about to be cut further

 

This is an unequal struggle. Ten times more images ( at great cost) to earn less money. 

 

Hence in 2012 I pushed the non-exclusive button and last year earned more elsewhere than at Alamy

 

What incentive to be image exclusive at Alamy?

 

 

 

Just to get my head around this. If being honest, is your portfolio as a whole of the same calibre/quality/good mix now, compared to 2006 or has it suffered in any way due to this enormous growth in ways such as overlap, similars, less time spent editing, keywording etc.? Are you seeing this "trend" elsewhere or is it just at Alamy where you're experiencing this?

 

I mean I agree, so does my numbers, that it has been tough and there is a great difference in profitability/returns 2006 vs now, but it doesn't look anywhere as dramatic as what I can deduce from your graphs. Forgive me, but I can't help but thinking "has he grown too fast?", "he might have let standards slip" or "there might be an awful lot of overlap/similars?". Not trying to be rude, I'm just being completely selfish and trying to get as much understanding out of whatever data one comes across - everyone's reality is different - mine in regards to Alamy is that it is one of the few places where I see pretty immediate correlation between work and sales, giving me reasons to think, despite the likely drop, where I get not the highest net RPI, but reasons to be positive and invest time. Yours look like you're working hard to just stay still, which can't be very motivating of course and makes me understand your upset much better.

 

Trust me, all expressed comments/thoughts are with the best of intentions, absolutely no malice or upmanship.

[Duncan_Andison]

3 minutes ago, M.Chapman said:

 

I'd like to see Alamy publicise the Alamy Rank algorithm to encourage the contributor behaviour they need (i.e. maximise revenue per image in their portfolio)

 

IMHO Alamy Rank should be based on

• a modified version of CTR% (which includes downloads and lightbox additions as well as zooms)
• and revenue (or unit sales) over the previous 6 months, divided by the number of images in a contributor's portfolio

If Alamy disclosed this part of the algorithm (in detail, for example 50% of your rank is based on.... etc.), and then made it clear how Alamy Rank affects your position in search results. This would encourage contributors to; 

• actively trim their portfolios to remove any "dross" and duplicates
• tightly keyword to improve CTR
• avoid uploading duplicates

At the moment the way Alamy Rank works is kept secret (for commercial reasons) and any info is largely "hearsay" on the forum. IMHO Alamy need to be more upfront.

 

Oh, and while Alamy are at it, remove the counter-productive "discoverability" index from AIM.

 

Mark

 

 

This will never happen, and rightly so. Once you publish how images are ranked it will result in gaming of the system. People will tinker, correctly or otherwise, images so they fit what is being looked for in the search. This results in each image having more or less the same rank and no there'll be no way to define which image should be at the top. 

 

For a search Algorithm, scorecard to work it has to be undisclosed and only then will it be affective. If you tell 100 people to choose a random number between 1-100 and that one of these numbers will be picked out for a prize, you'll get a randomised mix. If you tell them the winning number is likely to be between 1-5 then there won't be many selecting anything else other than those numbers. You'll either be splitting the winnings between the 100 or fighting over them. 

[M.Chapman]

10 hours ago, M.Chapman said:

We perhaps need to be careful not to assume increase in turnover has given a corresponding increase in profit. The increase in turnover may be largely due to a significant increase in the number of images, customers, and contributors. All of which increase costs. At the same time the revenue per image is falling, squeezing margins. A mere 2% growth in 2018 is very likely to have caused a fall in profits, hence the current panic...

 

Mark

 

Following up on my earlier comment. Alamy's operating profits (Turnover less cost of sales, distribution and  admin costs) were around £1.6M in 2017. Over previous 2 years their group staff costs have risen by about £1.7M. If their staff costs have risen again in 2018 then a 2% increase in sales may not generate enough to cover such increases. The weakening £ versus $ will also be having an effect. The weakening £ should make overseas sales more valuable, but Alamy said sales were only up 2%. At the same time overseas labour costs in USA and India (when converted back to £) will probably have increased. In the video James mentions that they monitor profit on a monthly basis. I  suspect Alamy's monthly profits are falling month by month and may even be showing a loss in recent months. Why else would Alamy announce a 20% cut on your contributors commission just before Xmas? Or is it just corporate greed after all?

 

Mark

 

[M.Chapman]

14 minutes ago, Duncan_Andison said:

 

This will never happen, and rightly so. Once you publish how images are ranked it will result in gaming of the system. People will tinker, correctly or otherwise, images so they fit what is being looked for in the search. This results in each image having more or less the same rank and no there'll be no way to define which image should be at the top. 

 

For a search Algorithm, scorecard to work it has to be undisclosed and only then will it be affective. If you tell 100 people to choose a random number between 1-100 and that one of these numbers will be picked out for a prize, you'll get a randomised mix. If you tell them the winning number is likely to be between 1-5 then there won't be many selecting anything else other than those numbers. You'll either be splitting the winnings between the 100 or fighting over them. 

 

I strongly disagree. If the algorithm is correctly designed the only "gaming" that will be possible will be to the Contributors and Alamy's advantage. The current mess doesn't help Alamy meet it's objectives. Where's the benefit in "random" behaviour?

 

Also I didn't suggest disclosing the full details of how images are ranked, only how a contributor's Alamy rank is calculated. 

 

Mark

[Duncan_Andison]

Just now, M.Chapman said:

 

I strongly disagree. If the algorithm is correctly designed the only "gaming" that will be possible will be to the Contributors and Alamy's advantage. The current mess doesn't help Alamy meet it's objectives. Where's the benefit in "random" behaviour?

 

Mark

 

Hi Mark. I used to work as a Credit Scoring Analyst for12 years before going full time photographer, designing scoring systems and applications. It's industry practice not to reveal this data as people will game the system... human nature. If you tell people what you want to hear, well, that's all your going to hear.

[Skyscraperfan]

If the algorithm just determines the ranking of the images in the search results, it is okay that the algorithm is hidden. However, when the algorithm is hidden, it makes no sense to reward somebody financially for getting a high rank determined by the algorithm. Just imagine a company which gives a bonus to its employees, but they have no idea why they got the bonus.

[M.Chapman]

22 minutes ago, Duncan_Andison said:

 If you tell people what you want to hear, well, that's all your going to hear.

 

I think you may have just contradicted yourself? Exactly, if Alamy want good behaviour, they have more publicly encourage it. The trick is having a ranking system without loopholes in it.

 

Mark

[Duncan_Andison]

10 minutes ago, M.Chapman said:

 

I think you may have just contradicted yourself? Exactly, if Alamy want good behaviour, they have more publicly encourage it. The trick is having a ranking system without loopholes in it.

 

Mark

 

No. The problem you're facing / confusing is the difference between search scorecard and search policy. Two entirely separate things. The scorecard will rank images based on their performance, history, age and other AI learning variables. Policy would be.... images need to be the correct size, right cameras, color space, must be keyword and have all required releases etc

 

Contributors need to know the policy but not the scorecard. Good behaviour will be following policy and it is not expected for contributors to understand or, need to learn all variables in a scorecard design.

[Skyscraperfan]

Doesn't the "Click Through Rate" already do quite a good job in that regard?

The CTR punishes keyword spamming for example. If you use keywords that do not really describe the image, your image will appear in the results, but people will not click on it. The same happens with low quality photos. Avoiding duplicates also increases your CTR.

My CTR usually is more than twice as high as the Alamy average. So it seems I might do something right.

[Martin Carlsson]

10 minutes ago, Skyscraperfan said:

Doesn't the "Click Through Rate" already do quite a good job in that regard?

The CTR punishes keyword spamming for example. If you use keywords that do not really describe the image, your image will appear in the results, but people will not click on it. The same happens with low quality photos. Avoiding duplicates also increases your CTR.

My CTR usually is more than twice as high as the Alamy average. So it seems I might do something right.

How's your conversion rate? Sales / zooms or sales / views.

[M.Chapman]

8 minutes ago, Skyscraperfan said:

Doesn't the "Click Through Rate" already do quite a good job in that regard?

The CTR punishes keyword spamming for example. If you use keywords that do not really describe the image, your image will appear in the results, but people will not click on it. The same happens with low quality photos. Avoiding duplicates also increases your CTR.

My CTR usually is more than twice as high as the Alamy average. So it seems I might do something right.

 

That's true, I just thought it would also be useful to add or publicise if Alamy Rank is also based on revenue or sales divided by the total number of images in one's portfolio to strengthen the incentives to not upload duplicate or lower quality images.

 

Mark

[Duncan_Andison]

13 minutes ago, Skyscraperfan said:

Doesn't the "Click Through Rate" already do quite a good job in that regard?

The CTR punishes keyword spamming for example. If you use keywords that do not really describe the image, your image will appear in the results, but people will not click on it. The same happens with low quality photos. Avoiding duplicates also increases your CTR.

My CTR usually is more than twice as high as the Alamy average. So it seems I might do something right.

 

I'd imagine CTR will be one of the variables in a search. As to what the weighting of that variable, only Alamy will know.

 

This is the thing. A lot of the variables in a search algorithm will be a numerical value. Image views - what % of images views of this image result in sales.... what % of the sales were editorial/commercial.... is the client an editorial / commercial client..... and so on and so on. To help get the right image in front of the client there will be many of these types of questions being asked that will result in the right image landing in front of the right client.... or at least, that is the goal. If contributors know the variables then they can influence the results rather than relying on the images previous performance / history as well as the clients buying patterns determining the final position.

 

Every contributor wants to be at the top of the search so there has to be way to determine the position of each image and this will inevitably be based on image performance and client needs/history... a scorecard will use a selection of variables that will do this form them.

 

[M.Chapman]

14 minutes ago, Duncan_Andison said:

 

No. The problem you're facing / confusing is the difference between search scorecard and search policy. Two entirely separate things. The scorecard will rank images based on their performance, history, age and other AI learning variables. Policy would be.... images need to be the correct size, right cameras, color space, must be keyword and have all required releases etc

 

Contributors need to know the policy but not the scorecard. Good behaviour will be following policy and it is not expected for contributors to understand or, need to learn all variables in a scorecard design.

 

I think we're sort of on the same page then... A key difference between a credit rating and Alamy Rank is that a Credit rating agency is presumably only interested in ensuring an accurate rating of "credit worthiness". They have no interest in improving the credit worthiness of members of the public.

 

Alamy Rank also needs to accurately rate the "quality or value" of a contributor. But it is in Alamy's interests encourage their contributors to improve their rating. Hence I believe it's in their interest to disclose what factors affect a contributor's Alamy Rank. 

 

Mark.

[Duncan_Andison]

Just now, M.Chapman said:

 

I think we're sort of on the same page then... A key difference between a credit rating and Alamy Rank is that a Credit rating agency is presumably only interested in ensuring an accurate rating of "credit worthiness". They have no interest in improving the credit worthiness of members of the public.

 

Alamy Rank also needs to accurately rate the "quality or value" of a contributor. But it is in Alamy's interests encourage their contributors to improve their rating. Hence I believe it's in their interest to disclose what factors affect a contributor's Alamy Rank. 

 

Mark.

 

Exactly, but that should be by submitting better quality images with better quality keyboarding etc rather than trying to workout % of this multiplied by this/ that and the other over time with a weighting of XYZ subject to client needs.

 

As photographers, we should focus on the product and how we keyword it. 

[Skyscraperfan]

13 minutes ago, Martin Carlsson said:

How's your conversion rate? Sales / zooms or sales / views.

 

Where can I see that? Alamy only shows me data from this year? And the "Your Images" section is even more limited. Wasn't it possible to see old data there? Now you can only go back to "last month".

[M.Chapman]

3 minutes ago, Duncan_Andison said:

 

Exactly, but that should be by submitting better quality images with better quality keyboarding etc rather than trying to workout % of this multiplied by this/ that and the other over time with a weighting of XYZ subject to client needs.

 

As photographers, we should focus on the product and how we keyword it. 

 

I still think it would be useful to give some measure of how important Alamy Rank is and what factors influence it (in percentages for example). If Alamy Rank only has a 1% effect (versus Tags, Captions, etc) then I wouldn't bother, but if it's a 50% effect I take notice...

 

Mark

[Duncan_Andison]

4 minutes ago, geogphotos said:

 

Only if there are sufficient incentives to do so and to cover costs. 

 

Haha... yeah, exactly. I try to work on a "zero cost to shoot" policy. Of course, I've invested in various pieces of software for motion/stills graphics etc but once they've been paid for, time is my main expense. A few props here and there and that's it. I would never pay to go to XYZ to just to take photos for stock. Takes too long to recoup costs before making a profit.

 

[Skyscraperfan]

My costs are tens of dollars per image and that does not even include the camera gear. The problem is that my subjects (skyscrapers) are far away from the place I live. So I already have the feeling that my images have to be more expensive because they were so expensive to produce. I would not suggest different prices for images though, but my high costs explain my unwillingness to accept another commission cut.

Perhaps I should switch to taking macro shots of fruits and vegetables. I could even eat those after the photoshoot :-)

[Duncan_Andison]

11 minutes ago, geogphotos said:

 

Isn't this the conundrum that Alamy and its contributors have to grapple with. Everybody says that they want high quality images - and photographers are keen to produce them - but nobody is prepared to pay properly for them! Cutting commission at a time of falling fees doesn't overcome this problem. 

 

Very much so. When it comes to illustrations, motion graphics etc I can produce imagery that is of a very high standard but after software, only cost is time. This of course is more of a commercial rather than editorial imagery though and, why for the most part I only submit some stuff here.

 

Edit... I think we maybe moving away from the original topic (regarding scorecards etc) I'd hate for this to be closed down 

[Martin Carlsson]

2 hours ago, Skyscraperfan said:

 

Where can I see that? Alamy only shows me data from this year? And the "Your Images" section is even more limited. Wasn't it possible to see old data there? Now you can only go back to "last month".

 

By keeping stats on your numbers over time using sales data and Alamy Measures. I do it by month by month - views, zooms, number of sales, total net sales values, portfolio size changes (how many I upload) etc. I do this pretty much everywhere where I license images and video - tells a lot, statistics is a powerful thing and works great as guide to ensure you spend your time right/wisely.