Some Alamy Statistics

[CLSI]

Let's get this out of the way first. There are three kinds of lies:  Lies, damn lies, and statistics. Ha ha, it's funny because it's true.

 

Also, yes, I am aware that the only statistic that really matters is the bottom line, the size of the check (cheque), the direct deposit, the number and size of the sales, what have you. Unfortunately, many of us don't get sales frequently enough to have them be statistically significant, and there's often such a long delay between anything that happens on the website (a view, a zoom, a change in rank, a change in the website) and the sales that may or may not result that it's nearly impossible to draw any meaningful conclusions on the basis of sales information, at least not until several months after the fact.

 

This thread is for those of us who actually enjoy looking for meaning in numbers. It is very unlikely to make anyone a better photographer or otherwise help to boost anyone's bottom line. In fact, to the extent that time spent reading this thread could otherwise be spent taking, processing, uploading, and keywording photos, it's more likely to have the opposite effect. All of which is to say, If you have a knee-jerk negative reaction to numbers in general and statistics in particular, there's no need to post about that here. Please feel free to move along to another thread; let us nerds wallow in our nerdity.

 

That said, being the nerd that I am, I have been keeping track of some Alamy statistics for quite a while now, and I thought it might be fun to share some of my findings. Initially, I did this in response to a heartbreaking drop in everything (views, zooms, and sales) for me around July 2010, and I wanted to figure out if it was just me, or if Alamy in general was seeing a similar slowdown. At the time, the best indicator I could find for customer activity was the number of pages of search terms that appeared in All of Alamy when no term was entered in the search box (essentially, this is what you see when you first open All of Alamy). I went back and determined this for 30-day periods starting in August 2009 (the earliest available month in AoA at the time), and I have kept this up since. Here's a graph of the results:

 

 

A couple of things jump out right away: First, December is clearly and consistently a slow month, and second, the overall trend has been upwards. (It became clear to me right away that I couldn't attribute my own drop in activity to any corresponding drop in Alamy activity more generally. My best guess now is that my own small collection--only about 300 images at the time--was of interest to only a handful of customers, one or two of whom retired or moved on to other jobs/projects in early 2010.)

 

Less obvious, but I believe (without actually having crunched the numbers) statistically significant, is an increase in activity following the launch of the new website in March. It's a little hard to tell for sure, because the overall trend is upward, and I'll admit I'm not quite nerdy enough to know which statistical tools could determine whether the apparent effect is real or not, but April, May, June, and July were all far ahead of any month that came before. August saw a slip back towards a number that might have been expected had there been no new website activity. What will September bring? Stay tooned...

 

At this point, it's fair to ask how good a measure of customer activity the "number of pages in AoA" is. After all, a thousand customers all searching for the identical search term will generate only a single line in AoA. More significantly, since we don't know how the folks at Alamy determine whose searches get included in AoA, the overall trend could conceivably just represent more customers being arbitrarily added to the list. A recent comment in this forum by someone official at Alamy suggests, though, that they define this set of customers in a way that will give meaningful statistics; in the absence of evidence to the contrary, I don't see that we have much choice but to take their word for it. As for customers coincidentally all searching on identical terms, I see no reason to expect this to happen more often some months than others. It may not be perfect, but I suspect Number of Pages in AoA is generally a reasonable proxy for customer activity.

 

Assuming that more searches lead to more sales, then, the pie is growing. Is your piece of the pie growing along with it? With the 40 millionth image in mind, it's worth taking a look at the number of images at Alamy over (roughly) the same period:

 

 

I didn't start keeping track of this until relatively recently, but I've determined that there were 17 million images in January 2010, and we hit 20 million in October of that year. I've interpolated the rest of the data for 2010, assuming roughly linear growth. Since 2011, the total number of images at Alamy has appeared on the front page (I used the Wayback Machine to dig up some of the earlier numbers).

 

The pie may be increasing, but the number of images competing for a piece of it is as well. To determine which is growing faster, we can look at the ratio of the number of images to the number of searches for each month. There are 20 searches to a page in AoA, so I multiplied the number of pages by 20 to get the total number of search terms for each month, and then divided the total number of images by that number to get the following graph:

 

 

From the point of view of an image sitting around hoping to be licensed, UP on this graph represents more competition per search term, and DOWN (obviously) less. Ignoring those December spikes (fewer searches increases the competition for each search), it looks like the situation was pretty stable--perhaps even improving--for much of 2010, then rose gradually, and has more or less stabilized again since September of last year (around the time the new buttons were introduced--coincidence?). In general, the number of images appears to have risen faster than customer activity, confirming (as if we needed confirmation) that we're in a Red Queen situation: we photographers need to keep running just to stay in place. And of course this doesn't take into account the number of images available from sources other than Alamy.

 

I recently learned (thanks to another member of this forum) that by limiting AoA to a week instead of a month it's possible to download data as an Excel file (I guess a month's worth of data is just too much to handle), so I've been keeping track of UCO, zooms, views, and even sales data more recently, but I think this post is already more than long enough. Any thoughts? Anyone see patterns or trends that I'm missing?

 

David

[NYCat]

I'm just fascinated by your avatar. What IS it? I've seen improvement personally but I think it is probably just a reflection of my own improvement in photography, processing, keywording. At least I'm not falling backwards. So far.

 

Paulette

[Phil Robinson]

A) It's the head and compound eye of a fly - and very good too.

That's interesting stuff, but a graph of Alamy turnover would be the real clincher. The best statistic we have for measuring trends is our own income per image, so it would be very interesting to see if our graph corresponds to Alamy's.

 

Thanks for the work, though, the trends are useful to know.

 

(Edit - the sarcastic face was supposed to be a 'B' followed by a ' ) ' but was corrected by the automatic face-checker.)

 

(Edit 2 - check out David's amazing chicken photos!)

[Chuck Nacke]

Thanks David,

 

Your presentation was worth the time it took to read.

 

I also appreciate that you use the term "license" and

not sell.

 

Chuck

[NYCat]

LOVE the chickens. Chickens can be amazing.

 

Paulette

[CLSI]

Well spotted, Phil, and thanks Paulette and Phil for the comments on my photos.

 

 

I also appreciate that you use the term "license" and
not sell.

 

 

You've trained us well, Chuck! (Some of us, anyway.)

[John Mitchell]

Well spotted, Phil, and thanks Paulette and Phil for the comments on my photos.

 

 

 

I also appreciate that you use the term "license" and

not sell.

 

 

 

You've trained us well, Chuck! (Some of us, anyway.)

 

I'm sold, but not totally on statistics. They usually tell only part of the story. Impressive number-crunching, though. Thanks.

[Duncan_Andison]

Great stuff. I've been working on a number of packages following trends with other agencies of late and due to move on to the zooms / views of Alamy. Interesting to see the steady increase in pages of search results each month, positive. 

 

I took a look at the data specifically to my account. My initial findings suggested that it would be tricky to tie zooms to sales. I'm going to sit down a draw up a list of things that could be tracked using the data they provide.

 

I've found the data collected for other agencies really useful for targeting subject matter and best agencies etc. Will follow / add to this thread over time!

[Paulstw]

Fantastic write-up. Thanks for taking the time. Your next job, the worlds economy  

 

I jest on the last part of course, but in the event of you wondering "Was it really worth it" then as a newbie I'd say this was valuable information. 

 

Sometimes the numbers don't lie, however, I bet you'd love to have been given better data to work with. I don't doubt AoA is useful and I've used it many times, but I heard that it's only figures based on registered customers, so the numbers may be more positive or negative based on a truer account of whose searching and buying. That info, I imagine, will be kept under lock and key. 

[Julesimages]

Excellent stuff and thanks for sharing. I'm no statistician but what you say seems to make sense. 

[DavidC]

I would love a synopsis of what this means in really simple terms for those of us who are only photographers - my concerns about all of the analysis that goes on is that it can only be as good as the figures which are available and

Alamy's figures are based on selected samples of registered customers, and as you know, when you start being selective, skewing can be introduced - this may of course be accidental or intentional.

 

I am truly impressed with the time and trouble that you have all taken, but I stick to my own analysis of what is happening to my own accounts - and put simply that is pretty dire. I have no intention of 'running just to stay in place' by upping my contribution rate - rather preferring the targeted response, to producing a large numbers of 'maybe' images further diluting a massively oversubscribed pool. I shall not be visiting tourist 'hot-spots' and asking other contributors for advice on the must-see spots.........

 

Sorry to intrude (you ask us not to) but I'm not totally opposed to nerdery and figure loving - and it is an open forum after all......

[Alex Todd]

Thanks for the work you put in and it is interesting although not precise due to all figures not being available to you.

 

What I would get out of the trends is that Alamy is increasing it's page views and therefore probably it's sales but as individual photographers we are getting less (according to many of the forum posts).

 

Like David I don't see the point of increasing my work rate to stay still. Think smarter rather than work harder

[CLSI]

 

I would love a synopsis of what this means in really simple terms for those of us who are only photographers - my concerns about all of the analysis that goes on is that it can only be as good as the figures which are available and

Alamy's figures are based on selected samples of registered customers, and as you know, when you start being selective, skewing can be introduced - this may of course be accidental or intentional.

...

 

Sorry to intrude (you ask us not to)

 

I don't see any reason to apologize--a healthy skepticism is important whenever one encounters statistics, and you (and Alex, among others) are right: there are some potentially significant unknowns in the available data. My "analysis" (if you can call it that) seems to confirm what most of us suspect, but if anything, that warrants greater skepticism: they don't call it confirmation bias for nothing.

[losdemas]

Love a good stats trawl: meaningless or not, I Iove a good few graphs to be sceptical about!  Thanks for all your efforts.  All interesting stuff: not sure if I could come up with anything meaningful though - beyond what you have already said, without making wild guesses and assumptions.  Shame that we can't have the full figures, eh?  But it would probably not be a good idea to be throwing that info around publicly, eh?!

 

Keep up the good work - look forward to future updates.