You mentioned Jack Keane... did the 3rd run really drop to 25? I suspect even that was too many! :)

When I originally posted this analysis, I did so with the stated and emphasized caveat that I believed there were errors in the data. I hoped this would cause those who monitored any portion of the sale I missed to come forward and provide the data they collected so that these tables could be as accurate as possible. Luckily, this very thing happened when Momo1991 provided me with a link and informed me that Mondo84 had recently posted totaled figures from his own data, and these differed from my own.

I contacted Mondo84 immediately, and worked with him to sort out the correct data from our two distinct databases. I felt that most of Mondo's 2nd & 3rd-run data - much of it personally observed - should be justly considered over my own mostly extrapolated data, and it was relatively easy to determine correct figures. Thus, though the updated data may yet contain errors, there's no longer any reason to think any of it remains incorrect.

As I posted earlier, I updated the tables several hours ago, but have just now updated the original post to reflect those changes, along with a plethora of other updates and corrections. For the sake of transparency, I've created a changelog for each, which you may view below.


Significant changes to the tables:

- "The Witcher 2" dropped from $6 to $5 for the 2nd run on, not just the 3rd.

- 2nd-run columns were added and populated with data. 10 deviations from the 2nd-run's standard "2x" pattern were factored.

- 14 3rd-run figures were corrected. This was the primary reason that the total number of games sold dropped by ~7,000 games, from more than 131,000 to 124,155.

- The amount of total sales was adjusted down by more than $30,000 to $360,616, from ~$392,000.


Changes to the original post:

- Any text regarding the 2nd-run info was updated to go along with the added 2nd-run columns.

- My description of how I rounded the prices is now much more accurate and specific.

- Total price is now always based on the lowest price offered for any given game, and not on first-run prices.

- The average price per game was corrected from "$3" to "$2.90".

- Any uncertainty was expunged from my since-confirmed conclusions regarding "Neverwinter Nights 2".

- My conclusion that "Leisure Suit Larry" did exceptionally well - better than 90% of the games offered - was retracted, as the corrected 3rd-run sales numbers were significantly less than previously believed. Granted, it still did extremely well in the promo (19th out of 101, better than 80%), and was 4th among games $2.50 or less. In its place were noted the top-dogs in the very-low-price category.

- Net GOG income was updated based on the newly corrected data, and to reflect WorldDan's point on GOG's 100% profit on all net sales of "The Witcher 2".

CDProjekt was one of the Warsaw stock exchange darlings some weeks or months back due to relative growth if I recall. It's definitively talked about but probably still quite small in absolute terms versus telecom, manufacturing, what have you.

GOG itself no. I don't know if the fin statements even discriminate CDP Red vs GOG vs any other wholly owned entity of CD Projekt.

GOG income could be lower. Because they have the same price for all customer they say they cover the difference in price from their own pocket. From their Terms of Use: "All prices include VAT or Sales Tax (rate used at the date of purchase) which will be applied based on the Country where the buyer resides".

In Europe the VAT is about 19-25%. Let's take one of the worse cases: a VAT of 25% (Only Hungary has more than that).
In this example if a game is priced at 10$ then it can be splited like this 8$ + 25% VAT (2$) = 10$. So, if VAT is 25% from the final price VAT represents 20% or 2$.

In the larger European countries (Germany, France, UK) VAT is only 19-20%, so from the final price VAT represents about 15%.

Instead of GOG making anywhere from 20% to 25% of the gross sales from this promotion they really make only 5%-10% or less from European buyers.

You know what? I could see GOG taking a tiny transaction fee hit. I CAN NOT see GOG voluntarily taking an entire VAT hit. Which makes me wake up and realize that they're not going to voluntarily take a transaction fee either. They're based in Cyprus - not Poland - for good reason.

Conclusion: few publishers, if any, are given gross points with GOG. I don't know why I assumed they were in the first place, frankly; it very seldom works that way around here.

Here's how it probably works instead:

Transaction fees and taxes are taken off the top. For sales to my neck of the woods, I can't see GOG paying much in sales tax - not from a country with all kinds of favorable tax laws (Cyprus) to a country with all kinds of favorable tax loopholes (the US), especially for online purchases. Whatever the case, the net dollars remaining are distributed 70/30. This means that GOG should effectively take only 30% of the hit from any and all charges.

Exceptions:
A. Server fees. I probably should have made this clearer before, but the additional server fees I added are a completely separate expenditure, not directly related to any sales figures. GOG's bill should hypothetically go up as a direct result of this sale, but it could be difficult even for them to calculate the exact amount. It's a nebulous amount and should not be (though this kind of thing often is) factored into a calculation of net income. That said, making an educated guess and tacking it on to GOG's final income may be prudent. One thing's for certain: you can certainly craft a contract that charges the publishers for your CDN service (similar crap is often pulled between writers and studios), but GOG doesn't strike me as that company. They pay the bill every month; they take the hit.

B. Free games. For a free game, there should never be a transaction fee or tax. GOG likely goes about these transfers in any of three ways, possibly only ever using one method, but it may depend on the situation.
1.) GOG simply pays the net amount the publisher would have received anyway. For a free game during a promo, this should be based on the discounted amount.
2.) GOG pays a flat, immutable royalty, not based on any price.
3.) The publisher agrees to release a set number of copies for the promotional benefit of everyone involved.

The bottom line is that GOG should still make between 21% and 28% every time, even with a high tax charge and a very high 5% transaction fee. After adding unusually high server fees due to a promotion such as this, one could make the argument that taxes - something I did not originally consider for online sales - could help push GOG's take below that range in some rare situations, but I doubt that's common enough to bother factoring.

Hey! I wonder if free games could be considered to tax deduction, in GOG case, as if they giving money to charity?

Kudos for these fine statistics!

they are not based in Cyprus, GOG is in Warsaw, Poland and if I recall correctly they actually share the same building as CD Projekt RED. It's there bank that is based in Cyprus.

Their physical headquarters is indeed in Warsaw, in the same building just across the entrance hall from CDP Red, yes. However, as recently as September 9th, 2013, for tax purposes GOG, Ltd. is registered as a corporate entity in Nicosia, Cyprus (scroll to the bottom). Thus, they are indeed based in Cyprus.

And as far as I know, they have been based out of Cyprus for virtually their entire history, though earlier I believe they were based in Larnaca, Cyprus, and moved to the capital, Nicosia, at some point within the last couple years.

This is cool data.

Thanks for keeping track of this and letting us see it. Interestingly, GoG knew the result (revenue) before going into the sale, since the design of the promotion (selling out all copies of each game in each round before moving to the next title) required that they make exactly that amount of money. No more, no less (except maybe for the final title). GoG decided the price point and the units to offer to sell. The only variable was the time it would take to generate that total revenue.

In most promotions, where a number of titles would be offered all at the same time and left to the consumers to pick and choose, there is more uncertainty as to the revenue that would be generated. Sure, there's probably price elasticity models that suggest what the velocity of a title would be at certain price points, but they take on a larger risk with that type of promotion.

Most important for me would be the velocity of the titles - how quickly was that revenue earned by title (units per minute, $ per minute, profit per minute) to compare each title's popularity and profitability. To look at each price point to understand how velocity was impacted by each threshold. To look at genres to see what was more responsive.

Then there are the intangibles that are really cool that they probably weren't expecting - the strengthening of the community, the new community members drawn in, a heightened sense of goodwill towards GoG for letting us join into the experience (and this promotion was as much as an experience as it was a promotion).

All in all, there's a lot of cool learning that can be done from this. Thanks for creating this and sharing!

I agree, I would have thought they would have made a lot more on such a popular sale...

$100,000 per week for a whole year is only $5 million. I guess my expectations are a bit off as well...

Edit: Also, Thanks ELI!!!

But what effort. Most of the effort was on our part, not trying to fall asleep, clicking really fast after 5 hours of Keane,et cetera.

+1 one for you. I agree with this observation. I've been GOG member since 2010 and most of the time I'd snap up a sale item really quickly and be on my way. I'd only jump to a specific game forum if I had a problem.

This sale changed my view of the 'General' forum area. During the waits for titles I wanted and the freebies, I spent an inordinate amount of time in the massive Insomnia sale forum. Seeing as there were many people giving away gift codes - I decided what the &@#$, and started to give away a bunch as well to be a part of the festive mood going on. The good karma was great to see from all participants!

I know about the publicity and they did get a whole lot of it during this sale. But again, thats partially why I would have thought more money would've been made. The insomnia thread was huge and there were people all over this website for all 5-6 days of the sale. So $100,000 seems a bit low for all that, $20,000 per day.
Then again, we don't know the true numbers as this thread is an estimation. Or maybe the sales were too good and not meant to make money but as jamotide said, only meant to pull in people.

Jack Keane demands a recount.

Don't Starve made some serious dough and sold quite a lot of copies.