jshacks, given that Valve won't obviously share sale data for their ongoing 'no daily deals' experiment, maybe we could infer something from Enchanted Steam server load?
You wrote that day one of the sale blew straight through all your expectations, how were days 2 and onwards in comparison?
Obviously, Enhanced Steam server traffic doesn't necessarily correlate into sales, but I'm happy to help provide any insight I can.
Wednesday was huge. Like, really huge. I know that Valve usually doesn't want to let their sales dates slip out (since it will discourage people from buying games on the run up to the sale) but it was pretty well known this time around when the sale was going to start. People may have been buying normally during the run up to the sale simply because refunds are a thing now, who knows. I think that because a lot of people (common folk who don't follow Steam much) knew of, and were anticipating a sale, that the numbers were higher than ever. Anyway, the first day of the sale shattered all Enhanced Steam server records in pretty much all respects besides number of downloads (more on that in a minute).
Subsequent days of the sale, despite having traditional "daily deals" have been pretty massive, but are more or less on par with what I'm used to seeing during a typical sale day. The numbers are higher than average for me, but I've also had about 300,000 Enhanced Steam downloads since the Summer Sale earlier this year. It's hard to say for certain, but since that's about 8% of the total install base and I'm seeing about 11% higher server loads, it's pretty straight across though slightly higher than the loads during the Summer Sale. Day two had about 23% of the total traffic as day one did. Subsequent days have slid about 10% per day so far, meaning that each day has about 90% of the traffic as the previous day. This is a little sharper decline than we're used to seeing with the sales, presumably because people kept revisiting the sales in previous years but also (again, presumably) because people this year have figured out that the sales aren't changing.
Another variable in the mix is that in all the previous sales for the past 2 years or so, Enhanced Steam has been mentioned in a major news article from a popular gaming website thus driving more traffic to the site, and subsequently the API. This sale is the first sale since the Winter sale of 2013 where that didn't happen. So it's possible that these new users skewed the "normalized" results vs this year's findings.
Hopefully this info helps. The TL;DR version: traffic was massive the first day and has been declining each subsequent day - sharp at first then gradual every other day, but there are a lot of factors involved so it's hard to say for sure the driving factors behind the numbers.
