One of the questions that faced sellers of popular music in the 20th century was the question of how much physical product to produce. In an era where physical records, tapes and cd's were the lions share of sales, this was no small issue. A fair percentage of stories involving artists getting screwed by labels involves the mechanisms by which unsold physical products were accounted for against sales.
The importance of this age old question has diminished as digital sales have become an increasingly large percentage of sales. Digital album downloads are not limited by manufacturing constraints, and they are not subject to return. However, because they cost an essentially equal price as physical products, their sales are typically going to follow the patterns followed by physical products: big first week, diminished sales in the following weeks.
The great exception to the way digital album sales mirror physical album sale patterns is in the area of pre-sale. The idea of a physical product pre-sale is typically limited to the "packaging" of more-expensive-than-average products that include the album/cd with additional accoutrements. If you look back to the golden periods for sales of specific physical formats: CD's in the 90s, albums in the 70s and 80s, the idea of a lengthy "pre-sale" period where people would be plunking down money for a regular old compact disc or LP is ludicrous (and prior to the invention of the LP the purchase of an "album" worth of music involved purchase of multiple 78 discs.)
So the category of "normal album itunes album pre-sales" is both attractive to the seller (no physical product required!) AND is essentially a brand new stream of income independent of post-release albums sales whether physical or digital. Itunes is the only digital service where a pre-sale can occur, and they are present in 119 countries (you can internationalize a pre-order link via websites like georiot.com.)
It is no wonder that the advantages of a Itunes pre sale campaign outweigh those of other pre-release sales strategies, most notably the "TopSpin model" of the larger packages for a higher price. The use of a single link for all countries creates a perfect model for evaluating the effiicency of various marketing approaches: A "conversion" is worth whatever the label's cut of the pre-sale price is. Let's call it five bucks. A conversion is worth 5 bucks.
Once you've established the value of a conversion, you need to establish a "ratio" for how many of whatever action (clicks, views, shares) create a single conversion. On-line conversions are "average" at 1%- but this does not necessarily represent an average for every measurement of on-line attention. The largest audience for on line marketing of a new record would be the number of impressions available via online ad servicing. Keeping with the number five, let's say a campaign for an album release receives 500,000. So at a one percent conversion ratio, that is 5,000 clicks/views/shares/likes, etc. Some of those people may purchase and "convert" directly out of that 5,000, but it's far, far more likely that they will do something else like watch a video, listen to a song, like a post, click on a link, etc. So out of the 5000, some part buy directly (let's say 5 perent) and then some people then convert out of the remaining 4750- say another one percent, so you have 250 + 48, close to 300 "conversions" at 5 dollars each, creating sales of 1500 USD.
The illustration of conversion ratios here is meant to show how large the number of impressions must be to generate a relatively modest number of digital pre-orders- it bears some resemblance to sifting for gold in a river, but of course does not hurt your back or require you to pan for gold in the 19th century. If you can increase the conversion ratios at each step- so that out of 500,000 you have 10,000 conversions initially you can increase sales. OR, if you increase the total of number impressions with the same conversion ratio you can achieve the same increase.
The real issue is whether you can purchase the appropriate number of impressions without overspending. For an artist with no existing fan base, purchasing 500,000 impressions let alone achieving a one percent conversion rate is going to be difficult to impossible. If you don't have access to an Audience that will support a half million impression ad campaign online, you need a label that can or you need to grow your own Audience to where that becomes possible.