A bunch of average app revenue data… and why you should ignore it
--
By Mike Sonders
The web is chock-full of app developers (and would-be app developers) looking for average app revenue data.
Presumably, these developers are looking for data that will help them benchmark (or forecast) the performance of their own mobile apps.
In this post, first we’re going to examine actual average app revenue data for the U.S. mobile market and demonstrate why these figures can be incredibly misleading when you’re benchmarking your own app’s revenue performance.
Then we’ll walk you through a much more reliable approach to benchmarking your app that doesn’t rely on misleading competitive data.
(Quick side note: This figures in this post include in-app purchase revenue. Advertising revenue is excluded, but that doesn’t change any of the conclusions or takeaways.)
Average app revenue: The ways it can mislead you
Whether you’re benchmarking the revenue performance of an existing app or you’re trying to forecast the revenue of an app idea, using average mobile app revenue figures can be grossly misleading.
Averages can include apps that are nothing like yours
Let’s say you’ve launched a mobile game app and you’d like to benchmark its revenue. That is, you’d like to know how your game’s revenue is performing relative to other apps. Is it high? Middling? Low?
You could start by considering the average revenue for all apps in the iOS and Android app stores.
But most of the apps in the app stores are nothing like your app. They’re terrible points of comparison.
In fact, you can see in the chart below that average revenue per app can vary dramatically among app store categories. On average, for example, a game app can expect to make much more than an app from any other store category.
Clearly it doesn’t make any sense to compare the revenue for one type of app to a global revenue average that includes all types of apps.
So, what if–instead of benchmarking a target app against the average revenue of all apps, which include apps that are very dissimilar in their purpose, audience, and functionality–we focus only on average app revenue of the target app’s specific category?
Well, we’ll run into problems (or, more specifically: noisy, unhelpful data) there, too.
Within each app category, there are thousands and thousands of apps. So, even though we’ve narrowed our focus by looking at a specific category, we’re still looking at results for lots of apps that are dissimilar in many ways. (And hence, aren’t useful for comparison or benchmarking.)
Returning to our example of games, we can see that even within a specific category, some types of apps will, on average, perform much better than others.
Comparing your app’s income to the average app revenue of all the apps within a category won’t provide much confidence in the resulting benchmarks.
So let’s see what happens when we go a level deeper.
Average app revenues are very skewed by top-performing apps
Let’s say your mobile game is specifically an “action” game. Will examining the average revenue per app of just action games give you useful benchmarks?
In short: nope.
Take a look at the difference between the revenue of the top-performing mobile game (Clash of Clans) and the 10th-best game (War Dragons) in the Action subcategory.
As of late October, Clash of Clans makes around 45 times more than War Dragons on a daily basis!
And that’s just looking at the 15 games at the top of the Action subcategory revenue chart; there are many, many more Action titles on mobile.
In fact, if we were to create a revenue distribution bar graph for this post showing the revenue of top 500 Action games, the bars showing the revenue of most games would be too small to see:
Notice something strange in those figures? The average revenue figure is higher than the 80th percentile figure.
That means that if you expect your hypothetical game app to drive the average amount of app revenue typically seen by Action games, then you expect your game to perform better than the top 20 percent of Action game titles.
In other words, the average figure is actually quite exceptional.
It’s great to have aspirational goals, but does that seem realistic?
If you glance over all the graphs you’ve seen in this post so far, you might notice a (power law) pattern that explains this phenomenon: the apps at the very top of the revenue charts make so much money that they meaningfully skew the overall average upwards when looking at any large grouping of apps (e.g., iOS apps, apps in the same category or subcategory).
This phenomenon makes average app revenue figures very misleading, even when you’ve drilled down to the level of app subcategory (or lower!).
Let’s look at the revenue of the top 200 Health & Fitness apps to illustrate this point:
The majority (over 50%) of Health & Fitness apps do not make any money from in-app purchases. Yet you’ll find that the average daily revenue per app for this category is around $1,500!
That is, the average experience in no way reflects the median experience; i.e., the experience of the app developers in the 50th percentile, where 50% of apps make more money and 50% make less money.
This disparity is a direct result of the top apps making so much more money than the rest of the apps in the category that the average app revenue figure is skewed misleadingly high.
In this case, the top app–Sweat with Kala–makes about $45 thousand per day… which is about $45 thousand more per day than the median Health & Fitness app makes with in-app purchases:
By no means is Health & Fitness the only category with a median daily app revenue of $0.
Books, Business, Education, Entertainment, Finance, Music, Photo & Video, Social & Communication and others all have a median app revenue of $0 from in-app purchases.
And they all have non-zero average app revenues well into at least three figures.
By now, I hope you can see just how very misleading average app revenue numbers can be.
So, if not with average mobile app revenue figures, then how should you benchmark your app revenue against those of other apps?
How to benchmark app revenue… without using averages
Now that you know you shouldn’t use average app revenue figures to benchmark your app’s revenue performance, here’s a much more reliable approach you can take:
1. Benchmark against apps that are very similar to yours
The more similar an app is to yours, the better it is for purposes of benchmarking and comparison.
Look for apps that are similar to your app in terms of app category, purpose, audience, design, tone, functionality, and features.
The apps don’t have to be exactly like yours (since that’s likely impossible to find), but they should be very similar across several of these dimensions, always including category and purpose.
Ideally, you’ll want at least a handful of very-similar apps against which you can compare. This will provide a range of revenue targets for low, medium, and high scenarios (discussed in the next section).
Using a feature unique to SurveyMonkey Intelligence, you can speed up this process by easily identifying apps that share your app’s specific narrow vertical, whether the vertical is dating, messaging, news, or one of over 30 other types of apps.
2. Look at average app revenue per user
Between two similar apps, the app with many more users is generally more likely to make more revenue.
That is, it’s not really a fair comparison to compare revenue between apps that have significantly different numbers of users.
You could try to find apps that are similar to yours in all of the previously-mentioned dimensions plus in terms of number of users, but in many cases you’ll end up with a very small sample set of benchmark apps.
Instead, when benchmarking your app revenue against the revenues of apps that are similar to yours, use average revenue per user figures. This allows you to ignore the total users of the apps you’re comparing.
A common normalized measurement of revenue among mobile apps is average revenue per daily active user, or ARPDAU.
You can calculate ARPDAU by simply dividing daily revenue by daily active users (both available from SurveyMonkey Intelligence), or you can use the SurveyMonkey Intelligence report that provides ARPDAU data on thousands of apps:
3. Identify low, median, and high performance targets
When app developers are looking for “average app revenue” figures, they’re really looking for the number that represents middle-of-the-road performance; the number that indicates an app isn’t stellar, but it isn’t terrible, either.
That number is a pretty useful one to know; it tells you that your app could be doing worse (whew), but that there’s definitely room for improvement.
So identify the median app revenue per user among your group of apps that are very similar to yours. This will be the number where 50% of apps do better, and 50% do worse.
Once you’ve identified that “middle” revenue level for your app, take it a step further. Determine exactly what it means for your app to be making “high” and “low” revenue so that you know how much room for improvement you have (or how much trouble your app is in!).
But don’t use the very best and very worst apps among your very-similar apps as guideposts. Remember: many apps make zero income from in-app purchases, and the very best apps make revenue that’s light years ahead of all other apps’ revenues. These apps are will create unrealistic (or undesirable) revenue targets for your app.
Instead, determine the revenue per user figures for the 80th percentile and 20th percentile of your similar apps. You want a “high” goal that is aspirational but conceivably within reach, and a “low” goal that doesn’t allow you to slack off or set your team’s expectations too low.
(Again, using these specific percentile figures isn’t a strict rule; you can choose other high and low percentiles, but I recommend against shooting too high or too low while making sure you give yourself aspirational “stretch” goals.)
Simply export ARPDAU reports to CSV from SurveyMonkey Intelligence, and then use the “PERCENTILE” function in Excel.
Conclusion
There are a couple of big reasons why average app revenue figures are misleading when you’re trying to benchmark an app:
- The average revenue per app for a large group of apps (e.g., all apps in the app store or in an app store category or sub-category) will include revenue figures for many apps that are dissimilar to yours. This automatically undermines the usefulness of the average as a benchmark.
- The apps at the very top of the revenue charts make tremendously more money than all other apps. This skews average app revenue figures upwards to misleadingly-high levels, where–for many categories–the average figures are higher than the 80th-percentile figures.
So ignore average mobile app revenue numbers. Instead, make informed decisions for your app by following these steps:
- Benchmark only against apps that are very similar to yours
- Look at average app revenue per user
- Identify low, median, and high performance targets
This post originally appeared on November 2, 2016 on the blog of SurveyMonkey Intelligence, a provider of competitive intelligence for the mobile app industry.
