I have posted a considered analysis on the iStock forum, which I'll reproduce here:

I posted that whole theory back on page 3 of the thread, but I deleted it, because after a few minutes, I realized it made no sense at all.  The main thing that bothers me right now is that it wildly fluctuates the value of the imagery.  How can an XS image be worth $20 in royalties one day, and $.20 the next, just because the buyer didn't spend the rest of the subscription credits on the first day?

Hate to say it, but I agree with Sean on this one.  I'll be pleasantly surprised if I'm wrong though.

Geopappas

Once again wrong logic, if you are non exclusive then you are selling the images somewhere else, but you have to pay for the shooting costs only once

Huh!!!

What does this have to do with your original statement???

Well you contribute to 9 sites (judging by your links)

so you costs are are spread over these 9 sites and are tax deductible

First, your original statement was that IS doesn't make 80% profit (from non-exclusives).  I was just replying to the statement and trying to show that contributors don't make 20% profit either (since we have costs as well).  What people are trying to show is that IS is already making a lion's share of the profit.  Now with subscriptions, IS will make an even higher profit %, while the contributors profit % will go down.

Second, yes, I am not exclusive (never said I was).  I don't believe in putting my eggs in one basket (as the saying goes).

Third, as always, you can't just say that costs are lower because a person is non-exclusive.  There are costs associated with uploading images (especially the time involved).  If you look at some of the top contributors, they hire staff to upload, keyword, etc.  So uploading to multiple sites costs money.  In addition, each site also has its own quirks.  IS has its funky tagging/DA system, Fotolia wants keywords in order, most sites have their own proprietary categorization system.  Many people downsize for subscription sites (which takes extra time).  IS wants mostly non-edited/out-of-the-camera images.  DT wants edited/saturated images.  So you can't make the generalization that costs are spread out evenly among sites.  Costs actually go up when submitting to multiple sites.  The object is to actually make more from each site than you spend.  That is why I don't submit to the dozens (if not hundreds) of small sites.  The cost for submitting to those sites isn't outweighed by the profits made.

Why don't we just wait and see what the details are?  I don't see much point in guessing.

I think Hatman hit the model on the head.  Istock especially wants 3 things IMO, screwing contributers on individual sales is not one of them (it is bad business logic).

1) Retain exclusives, it is one of the primary things that sets them apart from other sites and solidifies their overall collection as the best.

2) Gain new exclusives by pilfering from other sites, SS especially.

3) Make themselves more attractive to buyers.  Same as disambiguation, this attracts buyers.

Consider why the sub model is touted to have a different customer base, though only partially true, there is truth to it.  For example, designer A is a freelance designer and buyer.  The sub model doesn't help him too much.  Designer B is part of a 5 man team at a small company.  The sub model helps him quite a bit. 

Forecasting, and efficient use of time are very important to businesses as well as accounting means for determining price structure.  As is right now when per photo purchases are made by the design firm, it has to accounted to an individual project as part of the project price.  Initial cost estimating for the project is more difficult somewhat as well as the project cost tracking.  Anything that takes more time raises the cost of the image to the firm more than simply the purchase price.  Accounting and estimating time is not cheap.  Then also consider the photo acquisition process.  Per photo sales transactions take longer to complete for the buyer than simply downloading on a subscription plan.  Again, time is money, and the downloader's time is worth more than the person that could be doing the paying.

Consider what happens in a with a subscription model.  The lowest paid person in the building, the administrative assistant, take the time each month to renew the subscription.  Single point source on the outgoing money flow.  For accounting purposes the subscription (and thus all images used) are considered business overhead.  No photo has to be accountable to an individual project, and # of photos needed need not be estimated for project pricing.  Each project is charged the standard company overhead, which the subscription is part of.  Any member of the design team can log into the company's account and download any image they want, no time has to be spent concluding a money transaction, and the designer no longer is spending any money which has to be accounted for.  This is a big time savings internally for a company which streamlines operations for them, business management 101.

Istock has no need to screw contributers to gouge more money from them, it does not make business sense, as the contributers are customers as well and vital to the operation of their business.  Istock can win on two fronts, days when the sub isn't used, and attracting more customers that otherwise wouldn't use them because the individual sales model is not an efficient use of their company's resources.

If they price the sub model correctly, a token amount less than the per photo prices to attract buyers (remember the main benefits are not the cheaper per photo prices to the customer, the reduction in acquisition and accounting manpower required is, the unused credits are insignificant compared to time savings across the board), yet enough that Istock still makes a daily profit if all credits are used every day, it is a step in the right direction for them.

Unlike SS, pricing it so there is a fixed daily revenue (for them) benefits them quite a bit.  They can forecast (which SS cannot do with any real accuracy), not necessarily forecast profit, but they could forecast overhead revenue.  One easy way to price it would be to have the company overhead percentage as Istocks daily take.  They make no profit if every customer uses their subscription every day.  This would mean that unused days are Istock's profit intake, and the photographers are given the amount over the overhead on days the account is used.  Istock makes money in the end, and it helps the contributers as well and IMO is very attractive. 

Consider a day that a company with the largest available subscription downloads an XS from you as their only download of the day.  IS gets their cut so they are happy, the photographer sees a few bucks as payment for an XS.  Immediately their mind would be changed about IS's model and many more people would consider switching to exclusive as pricing is now fair, the photographers get their fair cut of the subscription, not a small amount as price fixed by an underpriced subscription model (hopefully IS realizes that the time savings to the customer and not the per photo $$ savings is the attraction of the subscription model, so sees no need to underprice it, that the $$ is not the primary attraction of the model).

The end of the day calculations would not crash their servers, in fact quite the opposite, it is more efficient.  Real time polling and updating databases take a lot of processing effort, the calculations contained within the constant real time update are insignificant compared to the resources expended on polling and the actual updating.  An end of the night calc at off peak times only has one large data poll across the database and one large update, the calculations would not overly tax the resources (actually it could be piped to another processor separate for the primary server processors to calculate, then the calculated data resent back to the primary servers).  In fact there actually is very little updating of anything in real time on any site, the active polling required to maintain real time updating is an incredible server strain, everything is just updated frequently enough that it gives the impression of real time updating.

It is entirely possible that their model was well thought out (which things at IS appear to be) and takes all of photographers concerns into account, and in the end the way they design their model is revolutionary in this it is beneficial to the photographer, that their intentions are actually to help the photographer and not screw them, even if every single customer migrated to their sub model.

well, based on the alexa rankings for istock traffic in the last 3 years we're all going down the dumper there... their traffic is going down the wrong direction.

Yet it is still has about 3x as much traffic as any other microstock site.

Alexa's traffic rating comes from people who have installed the Alexa toolbar, its not an accurate reflection of traffic to sites more the sites visited by people who use a particular piece of software

Istock has no need to screw contributers to gouge more money from them, it does not make business sense, as the contributers are customers as well and vital to the operation of their business. 

....istock makes money in the end, and it helps the contributers as well and IMO is very attractive. 

....
The end of the day calculations would not crash their servers, in fact quite the opposite, it is more efficient. 
....

What business are you in? Of course it makes business sense. This is a corporation, the bottom line and margins are all that matter. Especially when you're being taken over by a private fund. I'll eat my shoe if at the end of the day contributors aren't making less and buyer's aren't getting less for the dollar (with iStock pocketing the difference on both ends).

Also please explain how this wouldn't be in increase in load on the servers. You're going from a simple per transaction dual entry into the database to a complicated calculation based on how many downloads by that client had been carried out in the day. Think about it. There will be a lot more queries and inserts necessary per transaction than before. Think about all the table locking issues in this parallel system that doesn't have downtime to only do the calculations. I'd be shocked if iStock's system can handle this.

Sorry guys, but I think that in 2 or 3 years everybody will understand what happened today whit IS ,3 months ago? whit StockXpert. That's when the agencies will make profits over my back and sending me 25 cents for my work. Hope I'm strongly wrong on this

Sorry guys, but I think that in 2 or 3 years everybody will understand what happened today whit IS ,3 months ago? whit StockXpert. That's when the agencies will make profits over my back and sending me 25 cents for my work. Hope I'm strongly wrong on this

No way. What just happenned (about 15 mins ago) was that iStock clarified the details of their intention to wipe out the subscription based competition. Anyone who makes most of their income at Shutterstock & iStock will now want to be exclusive at iStock. They now own the RF stock business. And probably from now on and into the future.

I always quite liked Shutterstock, so that's a pity. Dreamstime can rot in hell for all I care.

Unless or until someone comes up with a new business model, iStock is RF from now on.

I'll eat my shoe if at the end of the day contributors aren't making less and buyer's aren't getting less for the dollar (with iStock pocketing the difference on both ends).

You want fries with that?