Why Are Your Players (Non-) Paying?
Vasily Sabirov at the GameNode Meetup
The devtodev company is focused on developing an analytics solution to help game developers augment income generated by games. During his lecture delivered at the GameNode Meetup, the project’s Lead Analyst and Co-Founder Vasily Sabirov shared methods used to analyze and segment a paying audience, which are instrumental in leveraging the way games are monetized.
Today, we will be speaking about the methods applied to analyze paying players, drawing on the example of free2play games. Let’s delve into their driving motivations, particularities and patterns of their behavior. Once you comprehend, what exactly drives your paying audience, you can apply this practice towards the non-paying, and therefore boost your income.
RFM Analysis
This method allows you to segment paying users by three indicators: R — Recency (time of the last payment), F — Frequency (regularity of payments) and M — Monetary (amount of payments). Every paying player is ranked from 1 to 5, with 5 being the highest score. Based on the assessment results, one can identify 125 segments and highlight those who need a deeper dive.
For instance, segment 555 represents users who have purchased recently, who are frequent and lavish payers. These are your VIP customers, who bring grain to the mill, they should be nourished and cherished. Or, for instance, segment 511 — these are the new payers . They have made a small trial purchase (scored 1 for Frequency and Monetary), however, quite recently (scored 5 for Recency). This highlights, that these guys are just starting to develop their payment behavior. How should we approach them? You should encourage them to make further payments — maybe they could be entitled to special offers, second purchase discounts, etc .
I would also like to talk more specifically about segment 155. These players used to make regular and large payments, however, a long time ago (scored 1 for Recency). How to deal with them? They represent a group that should be targeted for return by all available means. Seek to understand why they have stopped their purchases and act on the results. Grant them bonuses, send them push notifications or welcome back incentives, maybe you could think of a specially targeted product, as this segment represents money flowing away from you right away. Thus, RFM analysis enables you to group users into many segments and understand their behavior patterns.
For example, if properly configured, your game’s analytics system can help trace users’ migration across segments. In particular, you can spot players who have transferred to a less profitable segment and respond to this with the use of incentives to bring these players back to optimal profitability.
Segmentation by Payment Amounts
This is another method to analyze your paying audience, which is also based on segmentation. By payment amounts, users are traditionally divided into three segments: minnows — those who make small regular payments, dolphins — those who make larger payments, and whales — those who pay a lot. In our system, we have also introduced two additional segments — grand whales and grand dolphins.
Each segment is characterized by a particular behavior.
Retention
Paying and non-paying users also differ in terms of retention. Paying users are likely to show higher retention and remain loyal to a product, as they have something to lose — having already invested money into the game, they expect to get certain emotional feedback.
Operating Systems
The statistics show, that Android users have lower minnow-to-dolphin and dolphin-to-whale conversion rates, as compared with those who opt for iOS. Therefore, despite enjoying more users, Android would have lower APRU. The value of conversion into paying users would be lower, respectively.
Game Genres
As to the distribution of whales, dolphins and minnows across game genres, minnows tend to prevail in action, simulator, trivia and racing games. Whales are mostly found in RPG, as well as trivia games. The latter tend to attract people who are prone to spend large amounts.
Income Structure
If we take a closer look at the audience composition, we will see that 50% accounts for minnows, 40% — for dolphins and only 10% — for whales. However, the same 10% of whales are responsible for circa 80% of total income. In fact, the major portion of income is generated by a small group of massive payers, while 50% of minnows account for only 1% of total income. This is absolutely normal in free2play games. According to the statistics, iOS and Android have a different share of whales, dolphins and minnows in the total revenue income. Android is dominated by minnows, while iOS — by dolphins, as well as a higher proportion of whales. Your goal with respect to minnows is to convert them into dolphins as soon as practicable, and your goal with respect to whales is to retain them by all means. As shown in the analysis, whales are the guys that generate the highest percentage of income for your game.
Segmentation by the Number of Payments
Users can also be segmented by the number of payments they make. As a rule, 95% of income is generated through subsequent payments, and only 5% — through first payments.
For example, the guys from Tapjoy analyzed applications for which total income exceeded one million dollars, to identify user behavior patterns behind such income.
The analysis highlighted, that with 1,000 paying users who perform three payments within 90 days, 84% of such applications earn one million dollars. Furthermore — in the event that at least 35% of paying users go on to make third payments, an application is likely to earn one million dollars.
As per the survey findings, the importance of subsequent payments can not be underestimated. Your primary goal therefore is to ensure that the first payment is followed by the second and third.
By what means?
Firstly, users should experience emotional feedback after making the first payment — so focus should be to ensure that they get the desired joy in return.
Secondly, the feeling of joy should wane with time to prompt users to pay still more. As an option, move them to a higher league. They will start winning due to acquiring some new weapons, but their win rate will inevitably go down and they will be willing to pay more to sustain the required emotions.
Segmentation by Time in a Game
Another way to analyze your audience is to structure income by the ‘age’ of players. As a rule, the bulk of income is attributed to players who have remained in a game from 6 to 12 months. In other words, the longer a user stays in a game, the more they contribute to the project’s total income. If you build a historical chart, reflecting several weeks or several versions of your project, you will see how differently various categories of paying users react to product changes.
I had a case where the project changes didn’t affect the overall number of paying users, however, after analyzing the chart we realized that we had embedded a long-term risk into the game — the number of new paying users increased, while the number of old paying users dropped. It is known that income generated by long-term users is more important than that generated by newcomers. That is why we had to reverse these changes.
It is understood that newcomers are likely to become loyal and long-term users, but there is no guarantee that they will yield as much income. By analyzing the composition of the paying audience by the amount of time invested, you can predict your income and how much the project will earn over a certain time span. In other words, we know how many one-day folks are now in the game, how much they pay, and based on this we can calculate the probability of players’ shifting from one segment to another.
Using this simple predictive model, simulation scenarios can be created: what if we release an update that could possibly harm the long-term users, how this would ultimately affect our income..? Such models can be built in Excel.
Comparing Paying and Non-Paying Behavior
This method of analysis consists of sorting players into paying newcomers, paying long-term players, non-paying newcomers and non-paying long-term players. Each segment needs a tailored approach. For instance, non-paying newcomers could be influenced through paywalls that would prompt them to pay. Non-paying long-term players should be converted into paying customers by literally forcing them to make their first small, tiny purchase . They will end up paying again and again.
According to research data, the probability is higher that users opt for the second payment, as opposed to the first payment. And third payment probability is even higher than the second, and so on. This shows that enticing the user to make that first payment is the most difficult step to take.
Case 1 — Subsequent Payments
Let’s assume that RFM analysis identified lots of one-time payers in your project. The first payment averages 6 dollars, while the fifth payment — 15 dollars. It is evident that users have to pay more and more. Besides, first payments account for 80% of income. In this case, the problem is clear — users don’t make further payments. They either don’t get the desired joy for the money invested, or they do, but are reluctant to pay for some other reason. This is a wide-spread situation in poorly balanced games, when a player, having paid an insignificant amount, can complete the game. They get bored quickly and therefore such scenarios should be avoided.
Case 2 — First Time User Experience
It is essential to gain insight into the way that paying users think. It is commonly known, that the first time user experience matters — the first session, a game tutorial… A user might understand little of the game, but they respond to its visual style, the clarity of rules… Even minor changes to the tutorial may have a major impact with regard to retention and income.
You should consider the user’s first-payment scenario. How do they feel when visiting the store for the first time, what prompts them to go there? When they enter the store, what do they see? What goods do they view and what do they end up buying? These are the factors that lead users to their first purchase. For example, one of our projects was reported to have low first-payment conversions. We investigated the matter and discovered that a newcomer has to choose among two hundred items! They know nothing about the game yet and are already offered rubies, emeralds, swords, shields, magic bubbles, what not… It is not clear what to buy and they are seized by the so-called choice paralysis, when it is easier to buy nothing, than to make a choice. Therefore, we reduced the range of items for newcomers to 6–7 positions, which are easier to choose from, and the conversion rate climbed. As users stayed longer in the game, we gradually unfolded the store to them for further purchases.
Bottom Line
If you analyze the behavior of your paying audience, and compare with the non-paying, you will realize how to encourage users to pay more. Ask yourselves these questions: Why do the paying pay? What do they pay for? When do they pay? How soon are they ready to pay? As concerns the paying audience composition, you should monitor the correlation between the number of paying users and the size of payment, and how it changes over time. Take a look at your project from the perspective of a user, who is on the verge of paying. This might give you certain insights.
The above text is based on Vasily Sabirov’s presentation delivered at the GameNode meetup themed ‘Game Monetization’. Follow the link to check out the full version of the presentation.
