[As posted on http://sassybot.com/blog/atum-traffic-analysis-1-year-later/]
It has been just over a year since I launched the Atum website (the first recorded data in Google Analytics is from the 8th of October 2012) and I feel that it's time to give some insight into what happened with Atum over the course of a year. For those of you who haven't played Atum yet, you can go over here to play the game. Let's start off with a graph showing the traffic that was directed to the website over the course of a year.
The graph above illustrates 3 major points, at the end of October 2012, early February 2013 and at the beginning of March 2013. Atum was submitted to IGF around the 17th of October, but it wasn't until the 24th of October when we saw the first website report on Atum. The first review came from IndieGames, followed by Rock, Paper, Shotgun and PCGamer. These three websites are the main contributors to spike 1 and 2.
Spike 1 (Oct 22 - Oct 29) consists of the three websites mentioned in the previous paragraph, sending 712, 1142 and 467 unique visitors respectively. In addition, Reddit added another 660 visits during those first few days. All of the hits combined bring this to a total of roughly 4800 unique visitors. Spike 2 (Oct 29 - Nov 4) is when we see more non-US websites pick up the game. A Polish website called Polygamia picked up Atum which resulted in 532 visitors. The Danish website Ekstra Bladet ranks 2nd during this spike, sending just over 300 visitors to the Atum website. This spike also includes roughly a dozen other small websites which picked up the game, bringing the total amount of visitors for spike #2 to just over 4000.
The 22nd of January 2013 was the day that the IGF student finalists were announced. We pretended to be working all day while our browsers were on auto-refresh. Finally, the page was up and I ran downstairs for a very modest celebration (seriously, we had 1 beer and then went straight to making the required arrangements).
This is (weirdly) when the traffic dies down a bit and stays at around 200 hits per day for a week or so. The next spike (#3) happened just over a week after the IGF student finalists announcement (Feb 2 - Feb 4, 2013) and is almost solely generated from a single French website, jeuxideo.com. Jeuxvideo (meaning videogames) made a video about Atum. Luckily we have some friends over in Paris who were willing to translate parts of the video so we could get some cool quotes. Jeuxvideo.com itself sent over 4100 visitors to our website during that time, and after that contributed to a large chunk of the visits each day. The total amount of visitors for spike 3 ends up at just over 7000 (6100 of those coming from France).
After jeuxvideo.com we faced another 3 weeks of steady traffic, which changed when Vsauce3 mentioned Atum on the 25th of February. Vsauce3 added a 26 second video fragment of Atum in a roundup video of interesting things, which sent almost 7000 visitors to our site. For such a short fragment I would say that is quite impressive. The total amount of visitors for the 23rd of February - 28th of February ends up being almost 10000.
Spike 6 also mainly consists of YouTube traffic. Considering 'youtube.com/watch' is all the referral information Google Analytics gives you, it's difficult to pinpoint where this traffic exactly came from. My best guess is that YouTube sent out the weekly email with video suggestions, which then made the traffic from the Vsauce3 video spike again. The combined visitors from the 1st of March - 4th of March is almost 12000. That's a lot of traffic from 26 seconds of video. From this point onwards there is a consistent amount of traffic coming in from YouTube.
The last spike I want to mention is the 28th of March, which is when Kotaku published an article about Atum. We're all avid readers of Kotaku articles, so this was an amazing event to us. However, it did not send over that much traffic (just over 700 visitors). This meant that for those days, YouTube was still sending over more traffic than a newly published article on Kotaku. Furthermore, only 4.5% of the readers actually checked out the Atum website (the Kotaku article generated 15593 pageviews). That's not a lot, and may (or may not) have something to do with the inclusion of a gameplay video at the start of the article.
It has now been 7 months since Marta and Thomas went to San Francisco and we're still seeing a steady amount of traffic coming to the Atum website. In fact, we're still getting anywhere between 100 and 350 visits per day. Most of those coming from YouTube, Google and direct hits. Cumulatively, this brings the total amount of visitors to just over 100k!
For a small company like ours, this is a pretty amazing feat and we are very proud that Atum has made it this far.
This segment takes a look at how the amount of visitors from different countries compare to one another. As expected, the US tops the chart, not surprising considering that most of the biggest game related news sites that covered Atum are based in the US. Additionally it is also a pretty darn big country. 2nd on the list is France whose visits are almost solely generated by jeuxvideo.com. After that there aren't a lot of surprises. The Netherlands is fairly high up in the list, but considering that we are located there we expected to generate a decent amount of traffic from there.
The next point of data I want to discuss is the length of each gameplay session. I have to add that we did not add any data mining into Atum, so the number time could be skewed or just plain wrong. The numbers below are generated by the time that Google Analytics records.
That's right, 3 minutes and 21 seconds. It takes Thomas (Lead Designer) roughly 12 minutes to finish the game. He designed the level layouts and can navigate through the entire game without ever getting stuck. 3 minutes is not very far into the game, which means that most of the people did not see all of the objects you could interact with in the game. It also implies that only a very small segment of the players actually got to see the end. Unfortunately we do not have any data collection embedded in Atum, which means that I cannot give you any accurate statistics to support that assumption.
The most time spent on the page (on average) was 18:22, on the 27th of December 2012. Which would be around the time that the IGF judges log on and play the game. This seems to be the amount of time it would take for someone to finish the game. The number starts dropping drastically after that (the next highest average time on a day is 7:36). The top 10 days (sorted by average time on page) have an average above 6 minutes per session.
The most interesting statistic we got out of the last year is that a lot of people do indeed not finish a game. We had all heard of the statistic before, but were very curious on how it would work on our own games. Apparently Atum is not different in that statistic. We're very curious on how Atum will compare to our next title, Small Bang Theory and will definitely write an analysis comparing both of the games in a couple of months.