How much money I made writing web games in Flash

In an earlier post, I announced the start of an ongoing experiment to see how much money can be made from advertising with web games written in Flash. There has recently been some talk about this subject, as the developer of the Flash game Desktop Tower Defense reportedly made profits that would amount to over $100,000 per year (if the game were making that much money for a whole year, which I don't think it has been). Encouraged by this result, I wrote several games and promoted and monetized them in various ways. In this post, I give an account of what I wrote, how I advertised and monetized it, and how much money I earned as a result.

The Games

First, lets look at the games that I wrote:

Cave Flyer 1.0 - this was the first game I wrote for this experiment. It is based on the simple concept of flying through an ever-shrinking cave in a ship with lots of vertical inertia. It took about 4 hours to write in ActionScript 3. The graphics are weak, and it doesn't have a high score list. Nonetheless, it is a fairly classic concept without many Flash competitors (I think because it would have been hard to write efficiently in ActionScript 2), and people seem to find the game play fairly enjoyable.

Magic Plant 1.0 - this started out as as simple animation, and then i tried to add a game mechanic to it, which turns out to be too easy to be interesting to most. On the other hand, it has a certain zen charm that is strangely appealing, and while some people hate it, some like it. It requires patience, at least. It took about 4 hours to make.

Space Miner 1.0 - after seeing people's reaction to the somewhat weak Magic Plant, I decided to try my luck with a game which had a fun mechanic and a reasonable amount of effort in the graphics. The gameplay is based on the classic crystal catastrophe, although it is somewhat less diverse. I think I enjoyed this game much more than most people who played it, as I understood how deep and insanely hard it gets by level 20, and was able to get that far. Most people seemed to get bored because the first levels were too easy. This one has an online high score list- my hope here is that it encourages people to play again and again to beat the high score. But sadly, I hold all of the top 10 scores. This game took about a 3 days to write.

Xonf 0.9 - this game is pretty sweet in my opinion, though very hard. It has a pooled particle system that avoids triggering garbage collection, which can be an issue in smooth animation with ActionScript 3. The gameplay is taken from Mars Matrix, a hardcore overhead shooter that consumed many hours of my college years, as it was on an arcade machine right next to my room. So I had been planning on writting a game with Mars Matrix mechanics for a while... another idea that has been in the works for a long time, and which I was happy to work in here, is the procedural generation of bodies and flight paths for the enemies. Golan Levin's walking things is the inspiration here- he used flash to generate random walking insects. It seemed to me that this was a natural way to construct critters for a game. Spore, the game by Will Wright, is heavily influenced by the same idea. This also has a high score list and other people beat my score. Some people really like this game, and others don't, often complaining that it is too hard. I rewrote this game from scratch 3 times to optimize the particle system, so it wound up taking about 2 weeks, on and off.

How many elements can you name in 10 minutes? - Seeking inspiration for an easy game to write that would be popular, I looked at the most dugg games in Digg's Web Games Section. How many states can you remember in 10 minutes and similar games were very popular, and seemed pretty simple, so I wrote my own variation on this theme. It took about 45 minutes to write.

Spreading the Word and Monetizing Page Views

Those are the games. Now lets look at how I distributed and monetized them.

First, all of the games are hosted on my site, in a section for original flash web games. This is just an index of all o fthe games, with thumbnails and links to a separate page for each game. On the page for each game, there are two Google Adsense skyscraper format ad blocks. For some of the pages I used only text, and others text or images (I think that image advertisements are more appropriate for games, than, say lectures on economics). I will go over the Adsense data in the next section.

To promote the links to the games on my site, I submitted each of them to Digg in the Playable Web Games section. I also bookmarked them on my delicious account.

In addition to hosting the games on my site, I uploaded them to Kongregate and GameGum, both of which are flash game portals that offer ad revenue sharing with developers. To direct some of the traffic to my games on these sites back to my own site, I put a link in the games themselves back to the index of webgames on laserpirate.com.

Results and Data: How Much Money Did I Make?

So, how much money did I make from these games? First lets look at the page impressions and other related data. Google's Adsense terms of use prohibit me from giving away specific information about # of impressions, click through rates, etc, so I will be vauge: For these games, I got somewhere between $1 and $10 of revenue from Adsense impressions. I might also note that GameGum's revenue sharing system (allegedly) works by showing your adsense ads some fraction of the time on their site, although I'm not convinced that it works. If so, I think that those adsense views are not included in this total.

I can also give data on page requests from my own server logs, but the reader should be aware that these do not directly correspond to Google impressions, because they include non-unique views, many of which came from me during developement- and probably for some other reasons too (so Google: please don't smite me, I'm not trying to break any terms of conditions).

For the months of July and August, the number of requests to my server for each of the games was:

The elements - 261
Cave flyer - 89
Magic plant - 83
Xonf -101
Space miner - 142

We can also look at viewing statistics from GameGum and Kongreate. The stats for GameGum are (in # of impressions):

The elements - 18,091
Cave flyer - 198
Magic plant - 83
Xonf -201
Space miner - 121

I am find the count for the elements game to be quite dubious. If you refresh on the GameGum site, it count as another impression, so maybe someone repeatedly refreshed. I don't think I got adsense impressions for that many views, which suggests that they were not unique.

Here's the data for Kongregate (they provide statistics in 'Game Plays' and 'Ad Impressions'; these listing are ad impressions):

The elements - 1169
Cave flyer - 1263
Magic plant - 545
Xonf -1021
Space miner - 821

This totals to 4819 impressions, for which Kongregate lists a revenue of $0.51. This means that that are paying about $0.10 per thousand ad impressions.

It is also interesting to look at the number of times each game was dugg:

The elements - 11
Cave flyer - 3
Magic plant - 1
Xonf -2
Space miner - 1

From these numbers, we can see that my recipe of "take a popular game on digg that is easy to write and clone it" seems to have worked pretty well for the elements game. Digg didn't show much love for the other games.

Now lets look at some overall totals:

Time spent writing games: a few hours a day over 3 weeks
Games written: 5

# of views of those games on my server (not to be confused with Google Ad impressions): 676
Amount of money earned with Google Adsense: between $1 and $10

# of view on GameGum: 18694
Amount of money made with GameGum: I'm not really sure. I looked at the HTML they generate to be sure that it serving my ads, but I can't see any evidense of them being served in the adsense reports.

# of views on Kongregate: 4819
Amount of money made with Kongregate: $0.51

Conclusion

As this data shows, it is entirely possible for Flash web games to return very little profit, compared to the amount of time it takes to write them (except for the elements game, which shows that copying whats popular and being lazy is effective). I certainly would have made more money if I had spent that time writing software for my employer. But if I spent 45 minutes writing the next Desktop Tower Defense, that might be another thing. It is worth considering that none of these games are truly polished like Lumox 2, so I am not at all surprised that some people found them mediocre. It would be interesting to see how this goes with a game that has excellent graphics, sound and gameplay. It takes a lot of work to write such games. Maybe more work than it is ecconomical to do. On the other hand, really high quality games are often sponsored for $100-$1000. I will probably still write a few more web games just for the fun of it, but I won't really expect to make any significant amount of money. In contrast, my shareware titles such as Lumox 2 and Ultragroovalicious have sold many copies, producing a very reasonable return on my time investment for writing them. Seth Godin said something like "it is better to try to sell a product that costs $100 to 10000 people than one that costs $1 to 1000000 people". We can view the web gaming market as something like trying to sell a product that costs $0 to infinity people.