In-app purchases (IAP) are a popular way of monetizing apps and games. Although IAP has been around for a while and are widely used, it’s not always reaping developers the rewards they deserve.
In this post I’m going to discuss some of the signs and signals in your IAP metrics that might suggest your IAP economy is suboptimal, and how these signals can help you identify opportunities to grow your business.
First, I’d like to start by reviewing the structure of daily revenue and how all the metrics which make up revenue interrelate. To do this I’m going to use what Google Play calls the ‘revenue tree’ or the ‘revenue funnel’.
The revenue tree shows how the core components of revenue fit together and feed upwards to generate daily revenue. Each metric in the tree can be calculated by multiplying the two metrics below it, except for the dotted line boxes where you add the metrics.
The core monetization metrics, those in the red boxes, are divided into two categories: daily buyer percentage and average revenue per paying user (ARPPU).
The daily buyer percentage is, in my view, the primary monetization metric. This is because it’s always better to go for breadth when monetizing users: it’s safer to try to monetize a larger % of your players , versus trying to generate more revenue from your high spenders alone.
In contrast, ARPPU — which is made up of average revenue per paying user, average transaction value, and transactions per buyer — are the secondary monetization metrics. They are important too, however, they measure your ability to get value from users who have already chosen to pay. The potential downside with chasing these metrics is that it’s easy to push players to, for example, lift your average revenue per paying user, but this could be at the expense of decreasing buyer conversion rates.
To help you find the right balance, I’m going to walk you through the revenue tree. As I go, I will introduce some approaches I use to monitor revenue performance that can help you understand how a game is doing. I will then look at buyer percentage and ways to think about increasing returning buyers, especially the frequency at which users are paying. Finally, I’ll discuss the secondary metrics and offer you some food for thought about ways to change your games to improve them.
How to approach topline revenue
One of the common ways of driving top line revenue is LiveOps, which have become an integral part of driving game businesses and stimulating demand. However, LiveOps aren’t monolithic, they can be thought of as supply side sales and demand side events, and the balance between them. Supply side sales are the injection of discounted assets into the game economy; the ways you give users additional value for each dollar they spend. Demand side events are things such as tournaments, weekend challenges, and competitive engagements; ways to get players to use their asset balances. The relation between these two is important to create a balanced and healthy game economy that drives strong revenue.
To show how balance can be created I’m going to use two examples of different ways LiveOps can be approached.
On the left you can see an example of a developer who runs large sales approximately once a month. While the sales are running they generate a significant amount of revenue. This is followed by weekly LiveOps, however, the sale revenue spikes trend down over the month until the next sale. The developer on the right runs LiveOps almost every day. This balances their supply side and their demand side, and creates a much tighter revenue loop. Here the approach is interesting because the developer is trying to ensure that every day is appealing: every day users are perceiving good value in the game, and are willing to open their wallets and pay.
So, what is the best strategy? Well, the short answer is that one isn’t inherently better than the other. The right choice depends on your development team, skill set, game, and audience. How frequently can you create and run LiveOps? Is your audience driven more by supply or demand incentives?
Monitoring revenue with ‘revenue heartbeat’
The best tool for monitoring LiveOps is something I call the revenue heartbeat. To visualize the revenue heartbeat, determine the minimum and the maximum revenue days each month and calculate the average daily revenue for the month. When plotted, you get a visualization similar to this:
The ideal you’re looking for has the maximum and minimum revenue days forming a tight band around the average. As your revenue grows, the difference between the maximum and minimum revenue days is maintained; the band remains a constant width. If you see all these things, you may ascertain that you’re stimulating the right demand on a daily basis, and your players are engaging heavily and consistently.
In contrast to the first revenue heartbeat illustration, the next one shows some examples of suboptimal patterns.
At points A and C the developer ran large sales that created a significant hangover, illustrated by the spike in maximum revenue followed by a drop in the average monthly revenue. So, when running sales it’s important to find ways to minimize the hangover — its duration and depth — to avoid revenue going net negative when compared to pre-sale.
By way of contrast, at point B, the developer was tightening up their economy. They didn’t run as many sales but stimulated demand. That approach was very effective at getting users to drain their asset balance down, to try stimulate a desire to pay.
Revenue heartbeat is a useful way to monitor and understand the performance of your game, get a good sense of the effectiveness of sales, and whether you’re running them too frequently.
Other measures for revenue
There are also more advanced ways to look at revenue. One I particularly like is the coefficient of variation in daily revenue over a month. This is useful for quantifying LiveOps performance because it gives you a measure of the variability or volatility of revenue in a month. It is calculated by determining the standard deviation in your daily revenue over a month, then divide this by the mean.
When looking at this coefficient for games in Google Play, it exhibits a strong correlation with revenue growth. The coefficient can therefore help you understand the potential for your game to grow revenue. It can also hint at the optimizations and value that I recommend you look for.
Looking at the top 250 IAP games on Google Play for January 2017 to January 2018 shows that the majority of games have a coefficient of variation below 39%: there was some slight volatility but it isn’t extreme. An interesting picture emerges when this data is narrowed down to look at the games that had month-over-month growth.
The games with the lowest coefficient of variation were much more likely to see growth: over 55% of games with a coefficient of variation between 10% and 39% saw monthly growth. In contrast, less than half the games with very high volatility saw growth.
Revenue volatility is therefore an important factor to consider: is it at a level that’s healthy for your game? If it isn’t, then you may need to optimize your LiveOps cadence to balance the demand and supply side events to decrease volatility.
If you’re running predictable weekly or monthly sales, it’s easy for users to see the pattern. They may then choose not to spend at other times, stocking up assets from the weekly or monthly sales. Even if they run short of assets, they may hold off on purchasing if they know a sale will start soon. This sale hangover and anticipation can lose you money.
A solution to this is what I refer to as ‘predictable unpredictability’: your players get used to finding that something good is happening when they sign in, but they cannot anticipate what it might be. The important thing is that, because players cannot predict what the “good thing” will be, they’re not going to change their behavior before it occurs. So, ask yourself how predictable your offers and events are, and whether your players can predict what you are doing. If you are predictable, I recommend figuring out how you can shake things up.
We find that revenue often follows the 80/20 rule: 80% of the revenue comes from the top 20% of payers. This is certainly true when looking across the entire IAP games ecosystem. However, this could be risky for revenue generation and probably isn’t sustainable in the long term. Interestingly, it’s also not what the best performing games are doing.
When we looked at data from the top 25 Google Play IAP games, only six got more than 80% of their revenue from the top 20% of payers. The majority were in the low 70s, three were in the 60s, and two in the 50s. So, most top games are generating revenue from a wider range of payers.
Extending the view to the top 100 games, you can see that the further down the top charts you go, the more reliant games are on their top 20% of payers.
This begs the question: How reliant are you on your top payers? If you are, what can you do to broaden you payer base?
First, you need to understand where your revenue is coming from. To do this, I take monthly ARPPU and the number of unique days users are paying in a month. I then slice these into deciles, or 10 buckets, the top 10% in the first bucket, the next 10% in the second bucket, and so on.
To continue the illustration, the data is emblematic of patterns frequently seen in top performing titles in the Action RPG genre.
In this example, people in the top decile paid the most, on average $579 a month. For the second best decile the average was $120 a month. This follows the 80/20 rule. As the revenue continues to drop off it becomes apparent that the bottom 50% of buyers paid less than $10 each.
Is this an opportunity for growth, or something to accept because revenue from the highest value payers is stong? Personally, because I’m a growth mindset person, I think it’s an opportunity.
Before I discuss how to capitalize on this opportunity, there is another factor to explore: the number of unique days on which the buyers paid.
As you can see, half of the payers only paid on one day. This is a trend across Google Play, with most IAP games having between 40 and 60% of monthly buyers pay on only one day. How easy would it be to get these people to pay on an extra day each month? It doesn’t seem like it should be that hard to provide them with a compelling value proposition that incentivizes them to open up their wallets on a second day.
So what’s the opportunity here, what should you be focusing on, and where can you can extract potentially the most value? The answer is to focus on those users who are only paying on one day.
Looking again at the example game, you can see that 80% of the buyers who paid on only one day paid less than $10. This is a huge opportunity because, if your game has 50% of its buyers who paid on only one day and 80% of them paid less than $10, imagine what could be achieved if you could convert half of them to paying an extra $5. This is a huge upside and, more importantly, it’s sustainable. How would you do it? There are several options: targeted offers, sales, and exceptionally compelling value propositions for these users.
Importantly, to improve your game’s performance, think about that value proposition. Consider the following:
- How you’re communicating value for money to users.
- How compelling the ‘value’ is.
- How easy it is for users to understand what they are getting.
Also, think about your sales segmentation and how you’re presenting it to users. Think about the psychology. Would a person who occasionally spends $5 and who then sees a $100 SKU in a big sale get excited about a $5 offer? Alternatively, would they see the significant value proposition being offered to high value spenders, think that their $5 won’t let them compete with the players spending $100, and as a result leave their wallet shut. So, consider how you’re segmenting your users and how you’re presenting the offers. Try to give your players offers that are targeted to their spending patterns and that will appeal to them.
Another useful tactic is daily deals, something that has become quite popular. The way the top games implement them involves good communication of the deal’s value, scarcity, and aspirational goals for making the purchase.
Finally, I’d like to mention an idea about repeat buyers bonuses. For example, after a user has made their first purchase in a month, you could offer them a 10% bonus if they make a second purchase within 7 days. This is a great way of encouraging players into a weekly purchasing habit.
Optimizing secondary metrics
Across the top 250 IAP games on Google Play, each day purchasing users make between 1.5 and 1.9 transactions, with an average transaction value of between $8.50 and $25. This results in an average revenue per paying users, depending on the game, of somewhere between $13.70 and $44.50.
These are significant amounts, but often they come from a very small set of users. It is therefore important to consider what the relationship is between these metrics and what opportunities exist for improvement.
There is a positive linear relationship between average revenue per paying user (ARPPU) and the average transaction value (ATV): if you can get users to pay at a higher price point, you’re likely to get more revenue from them. Also, given that the transactions per buyer on a daily basis are only 1.5 to 1.9, the primary driver for average revenue per paying user is the average transaction value.
In contrast, the plot of average revenue per paying user against transactions per buyer is just noise. Attempts at cluster analysis, curve fitting, and other similar analysis found no reliable pattern. So, this metric is probably tied to a game’s design and core economy, which means we might be able to leverage game design to bring more order to the chaos seen here.
If your game falls below the 25th percentile, you should ask the following questions:
- How many assets do you sell in your game?
- Do you sell only one asset (gems or coins or credits)?
- What price points are you selling at?
- If you have a low number of transactions per buyer, and your IAP sells at very high price points, what is the velocity of money?
- Have you created a scenario where users are incentivized to buy the largest pack possible because of bonuses you offer on top of them, resulting in users riding those assets for as long as possible? You know that they will spend them eventually, so there’s no urgency to buy multiple times.
At the other end , if you’re above the 75th percentile, you may want to ask the following:
- How many assets are you selling?
- If you’re selling two or more asset types, you’re in a very healthy range. If you’re only selling one kind of asset, what kind of impulse buying decision making are you creating in your game?
- Have you created a scenario where users are incentivized to make a large number of very small transactions each day?
If the answer to the last question is yes, then there’s the risk that you’re leaving money on the table because, every time a user has to decide about making a purchase there’s a chance that they’ll stop. To address this, look for opportunities to upsell to your users. Instead of showing them X gems for $2, show them a $2 and $5 offer and see if they are interested in the higher offer .
There are opportunities to optimize and improve here, but the right optimization and improvements are likely to be game specific. Finding the right approach depends on asking your development team the right questions:
- What is your velocity of money?
- How many impulse purchases do players make each day?
- Does your game incentivize asset hoarding?
- How is your game economy designed?
- Is it a capital expenditure economy where users are primarily purchasing assets, such as gems, and spending them to get permanent upgrades?
- Or, is the economy more focused on purchasing consumable assets, such as coins or credits, and players spending them as part of their daily play?
Looking for the signs and signals of suboptimal monetization is an effective way of driving revenue growth.
The top games on Google Play show that the lower the daily revenue volatility, the more likely games are to see revenue growth. So, assess volatility and ways to minimize it: look to see if your LiveOps are balancing supply and demand side events.
The top 25 games on Google Play rely less on their high value users compared to games lower down the top chart. Find out how closely your revenue follows the 80/20 rule and whether there is an opportunity to broaden and diversify your payer base. Look for ways to get users who spend once per month to spend a second time.
There is a strong linear correlation between average transaction value and average revenue per paying users, however the relationship does not exist between transactions per buyer and average revenue per paying user. The optimizations here are likely to be very game specific and an opportunity to exercise your creativity.
Addressing these three areas may help you improve your game revenue growth strategy and will hopefully lead to tangible results.
What do you think?
Do you have thoughts on these approaches to optimizing game revenue? Let us know in the comments below or tweet using #AskPlayDev and we’ll reply from @GooglePlayDev, where we regularly share news and tips on how to be successful on Google Play.