Please read the Disclaimer

10 days ago I launched a survey in the iOS game developer community aimed at gathering revenue data from as many developers as possible. The goal was to get a more realistic view of what the iOS game marketplace actually looks like and share the results with the entire dev community. The reason I felt this was necessary is that we tend to see two kinds of articles written about iOS game revenue: “Developer makes millions on iOS games!” or “Game makes $0 on App Store”. I felt it was important that we get a more realistic look at what the market we’re developing for looks like.

With that in mind, the survey launched on Monday, September 19, 2011 and ran for seven days. 252 developers filled it out! Now, before I get to the actual data, there’s some important disclaimers I need to make, and I want to talk about methodology. Let’s begin…

Methodology

The survey was conducted entirely using the online service SurveyMonkey. The survey consisted of eight questions. Two questions gathered information about the type of developer and the number of people working on their games. The next three questions gathered data about lifetime game releases on the App Store, and revenue those games generated. Finally, the last three questions gathered data on games released within the last 12 months (more on this in the next section on errors and bias).

Requests to take the survey were distributed via the following social networks and web sites:

• twitter
• facebook
• Google+
• 148Apps.biz
• Reddit
• TouchArcade forums
• iPhoneDevSDK forums
• cocos2d forums
• Unity forums

The goal was to engage as many active iOS game developers as possible. More on this in the next section on errors and bias.

The survey was closed on Monday, September 26, 2011 at noon, EDT. Survey responses were downloaded at that point, and I’ve been using Numbers (from iWork ’09) for Mac to analyze the data.

It is also worth noting that when I launched the survey, I stated two things:

1. The survey would collect no personal data
2. The data would only be released in aggregate, no raw data would be released

This is why I’m not releasing the raw data.

Errors and Bias

Disclaimer: I make no claims as to the statistical validity of this data. There is a good chance that the sample population is not representative of all game developers on the App Store. There is a good chance that I introduced measurement bias/error into the data by the way I worded the survey questions. In short: I am not a professional statistician. Do not treat this data as 100% accurate. It is just interesting to look at.

Ok, let’s talk about that disclaimer in a bit more detail.

Sample Population

Because the survey was completely voluntary, and I have no information on the demographics that make up App Store developers, I have no way to determine how representative the data is. Further, because of the way in which I went about gathering the data, the developers who responded are all likely developers actively working on games, and who are actively involved in the development community. Because of this, I would be tempted to guess that the numbers we see here are actually higher than on the App Store over all. However, I have no data to be able to back up that guess.

12 Month Data

One of the things I really wanted data on was a snapshot of what the last 12 months have looked like for game developers on the App Store. However, the questions I created to gather this data clearly confused respondents. The intent was that devs enter only revenue from games released within the last 12 months, but many developers provided revenue from the last 12 months for all their games. This made the data I collected for these questions largely useless for the purposes I wanted. Further, most people didn’t understand the instructions and didn’t match the sales revenue to the non-sales revenue in the two questions, making drawing conclusions there impossible, also. You’ll see later on that I did manage to get some basic data from it, but couldn’t do the detailed analysis I had hoped for.

Other

There are, no doubt, other sources of error and bias in the data. The main thing to remember is that these numbers are not 100% accurate, but rather just provide a glimpse into the App Store market for games.

To the data!

Results

The survey was open for exactly seven days and had 252 respondents.

General Questions

The first two questions of the survey were there to get an idea of the kinds of developers responding.

You can see from the results in Figure 1 that only about 1/3 of respondents consider themselves full-time independent game developers. Over half the respondents are part-time indies, hobbyists, or students. (Note: Click the charts to see them full-sized).

When questioned about the number of people (from now on referred to as “developers”) working on their games, half the respondents were working by themselves. I was surprised by the number of respondents who were working for larger companies in the 10+ developers range. However, over 93% of respondents have 5 or fewer developers working on their games. See Figure 2.

Lifetime Game Releases and Revenue

The next three questions of the survey gathered information about lifetime revenue on the store. The three questions asked for data on:

1. The number of games released on the App Store
2. The number of months the developer had games on the App Store
3. Lifetime revenue for all games on the App Store

Using this data, it’s possible to generate some very interesting results. First, let’s look at all the lifetime revenues reported (see Figure 3). One of the most interesting features of the graph is the clearly exponential curve associated with the revenues. The graph makes it very clear that most developers aren’t making a lot of money selling games on the App Store, while a few are making a lot of money.

I’ve made note of both the arithmetic mean average, and the median average on the chart. This is why the median is so important. The extremely high revenues reported by a small number of developer skew the arithmetic mean significantly. If you looked at that as an average, it would be easy to say “the average game developer has made about $165,000”. However, the median tells a very different story. The median splits the developers in half. This means that 50% of developers have made less than $3,000 lifetime revenue on the App Store, while 50% have made more. The reason that the mean and median are so different is that the computed sample standard deviation is 639,966. A standard deviation that high means that the mean average is not very representative of the data spread. Because of this, I have used median averages everywhere in these results, instead of mean averages.

What is also telling is that if you were in the 75th percentile, you would have made about $30,000 on the App Store. This means that only 25% of developers have made more than $30,000 lifetime total revenue selling games on the App Store. Conversely, we can see that 25% of developers have made less than $200.

I wanted to take that revenue data and plot a distribution curve from it. However, the range of data was so large that I couldn’t plot it on a linear scale. This was the case for many of the graphs you’ll see. Like Figure 4, I have made note whenever one of the axes is using a logarithmic scale instead of a linear scale. By breaking the revenue down into buckets (each one 10x greater than the last), I was able to get a better distribution graph (see Figure 4). From this, we can clearly see that nearly 25% of developers have made between $1,000 and $10,000 on the App Store. What is particularly impressive is that 4% of respondents (10 respondents) had made over $1,000,000 on the App Store!

Note on Figure 4: You’ll notice duplicate values between buckets (i.e. 1-10, 10-100, etc). This was done only for the labels so the chart was easier to read. The data is actually divided into (10ⁿ)-(10ⁿ⁺¹ – 1) buckets (i.e. 1-9, 10-99, etc).

That 4% of respondents got me wondering about the idea of where the revenue was going on the App Store. So next I took a look at what percentage of the total revenue reported was reported by what percentage of respondents (similar to a distribution of wealth chart you might see for a country’s population). See Figure 5. What is fascinating to me is that the top 20% of developers are earning 97% of the revenue on the App Store, with the top 1% earning over 1/3 of the revenue on the App Store. The bottom 80% of game developers are earning only 3% of the revenue.

Next, I wanted to start comparing lifetime revenue to the other data the respondents had provided. To start, I wanted to see how revenue compared to developer type. Figure 6 shows a graph of median lifetime revenue, divided by developer type. It is no surprise to me that full-time indies, and representatives for iOS game dev companies reported the highest revenue. Note that revenue in Figure 6 is charted on a logarithmic scale, so the median earnings of a full-time indie developer are reported to be 30x greater than those reported by part-time indies. This does not mean that going full-time indie will guarantee you 30x the revenue, these are just the numbers that have been reported.

Next I wanted to see if developers who worked with more people earned more revenue than those working alone. Figure 7 clearly shows this is the case. Note that the revenue for companies with 10+ employees may not accurately reflect a good median, because there were so few responses in these categories. However, it’s clear that individuals have earned the least lifetime revenue, on average.

But then I started to wonder if it was just because larger groups might be able to release more games, meaning their total revenue would be higher. So, I broke the revenue down, dividing it by the number of games released, and by the number of months the developer had had apps on the store. The result is a chart of median per-game, per-month, lifetime revenue by the number of developers who worked on the games. You’ll see in Figure 7b that the curve looks almost identical to Figure 7’s. The conclusion that I draw from this is that, in general, larger groups of developers are able to create games that earn more money. Wagering a guess, this is perhaps because they are able to create games that are larger in scope, more technically interesting, and more polished, because they have more people to work on the game and provide input into its improvement.

Finally, a friend was curious whether or not releasing more games meant that a developer ended up improving over time. To attempt to answer that question, I divided each respondent’s lifetime revenue by the number of games they had released on the App Store, then graphed the median distribution curve by the number of games released. The results can be seen in Figure 8. What is really interesting to me is that developers do seem to generate more revenue over time (on average). This should be encouraging if you really want to make games, but your first game was a flop. Fear not! 50% of developers who have only released one game made under $500 on that game. However, the more games developers had released, the more per-game average revenue they seem to generate. This seems to validate the old adage: practice makes better than doing something once. Wait…that’s not quite right…

12-Month Releases and Revenue

The final three questions of the survey were supposed to deal with revenue generated by apps released within the last 12 months. However, since many respondents provided the last 12 months of revenue from all their games, I can’t draw the same conclusions that I would have liked. However, what I have done is graph the revenue for each individual game reported.

Figure 9 shows the individual game revenues over the last 12 month period. It is a graph of 382 games that reported non-zero sales revenue (including IAP). You can see that the curve follows a very similar line to the lifetime revenue chart, in that it’s exponential. Again, we can see that the difference between mean and median is significant, telling us that the high earners on the right distort the mean average.

What’s important to note in Figure 9 is that the median game earned $1,100 in the last 12 months. This means 50% of the games earned less, and 50% earned more. 25% of games earned less than $140, while conversely, 25% earned more than $10,675.

And finally, Figure 10 shows the non-sales revenue generated by the 85 games that reported non-zero revenue. The non-sales revenue was to account for all revenue generated from a game aside from sales (e.g. ads, affiliate links, merchandise, etc). Figure 10 is graphed on the same y-axis as Figure 9. You can clearly see that the top end is much lower than for sales. However, in the middle of the graph, games that reported non-sales revenue, reported slightly higher earnings on the non-sales side of things.

What this means is that there is clearly some good revenue to be made through things like ads, affiliate links, and other non-sales sources of revenue, and there are clearly some games doing this very well.

Conclusions

Phew! Did you make it all the way through? Good. This is a lot of data to process, so thanks for following along. I’m sorry I wasn’t able to get more useful data out of the 12 month questions. I clearly needed to word the questions differently.

Hopefully these results have provided some insight into the games market on the App Store. I hope that it can be used to help set expectations for new and experienced developers alike. It is clear that there is a lot of money to be made in games on the App Store. However, as the data shows, it’s not easy, but the more games you make, the better you’ll get. Common sense, I suppose…but sometimes it’s nice to have the data to back it up.

Thank you so much to the 252 people who participated in the survey for sharing this data with the rest of the community. We all appreciate it greatly.

Now you’ve been reading this for much too long. Get back to work on your next game!

Owen

This entry was posted on Wednesday, September 28th, 2011 at 2:55 pm and is filed under Business, ipad, iPhone, Marketing, Misc. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.