The most common method for monetizing free-to-play games by far is selling lump sums of virtual currency in exchange for real money. But this revenue model comes with a major drawback: It does nothing to promote paying players’ re-engagement with a game. If they log out with 1,000 gold coins sitting in their account, then that same 1,000 gold coins will be sitting there whether they log in tomorrow, a month from now, or even years from now. And in the interim, it’s easy for players to move on to other free-to-play games, increasing the chances that they’ll never log in again.

So while lump sum sales of virtual currency is an effective monetization strategy, it could certainly use some help. An excellent complement, which I explore in this post, is the annuity model.

An annuity is just a sequence of payments that a person receives in exchange for an initial investment. Interest payments by a bank into a savings account are a common kind of annuity. Insurance companies usually offer another type of annuity where you pay a large sum now (say $100,000), and then receive a guaranteed amount – say $5,000 a year – for the rest of your life.

Annuities in free-to-play games work similarly: you pay a certain amount now in real money, and in return you get a guaranteed stream of daily payments in virtual currency over some specified time period.

While annuities are still relatively rare in Western games (they've been around for a long time in China), we’re starting to see some early, interesting examples:

FTP Annuity Case Study: NetEase’s Eternal Arena

NetEase’s Eternal Arena

With over 1 million downloads on Google Play alone, NetEase’s Eternal Arena is a popular mobile action-RPG with MOBA elements available for Android and iOS since November 2015. Its also partially monetized using an annuity model. Before discussing how the annuity works, here’s a quick overview of the game’s virtual economy:

In Eternal Arena, players control up to three avatars at a time as they play through a single-player storyline. (They can also participate in League of Legends-style PvP.) Players can choose amongst dozens of avatars and equip them with myriad weapons and armor, all of which they can obtain and upgrade by spending diamonds, the game’s premium currency.  

Eternal Arena draw

All acquisitions and upgrades are achieved through a randomized, gacha-like “draw” system. A single draw is free once every 24 hours, or can be purchased immediately with 198 diamonds. Alternatively, the player can purchase a “10x draw” for 1,982 diamonds, which guarantees a very high-end new avatar or piece of equipment. Small sums of diamonds can be gradually earned through various gameplay activities – logging in daily and the like – but to rapidly acquire and upgrade new avatars and items, players must directly purchase diamonds.

Eternal Arena’s annuity offers

This is where the annuity comes in. NetEase offers two different but complementary options for buying diamonds: a lump sum, or an annuity. There are two available types:

The two annuities can also be stacked on top of one another (though players cannot buy more than one kind of annuity at once). And of course, owning an annuity does not prevent players from buying as many diamonds in lump sums as they wish.

On a pure unit price comparison, either annuity option is a steal compared to any of the game’s lump sum options. In Eternal Arena, the diamond package with the best discount provides 14,000 diamonds for $99.99, an implied unit price of $.71 per 100 diamonds. By contrast, the 30-day annuity option comes out to $.15 per diamond, and the perpetual annuity (or perpetuity) has a theoretical unit price of approaching zero. This is a nice deal for the cost-conscious player. It’s also attractive to big spenders, who would likely pay for the additional stream of diamonds on top of the $100+ worth of diamonds that they’d likely buy over the lifetime of gameplay, anyway.

Just as important: Annuities are also great for boosting player retention, which is key for maximizing lifetime value [1]. No matter how long they’re away from the game, players who choose the perpetuity option for $9.99 will always have an incentive to return and use their accumulated diamonds. In fact, the longer they’re away, the more attractive returning becomes, because players know they’ll log in with a bounty of amassed diamonds they can spend to easily catch up with their old friends or engage with new content. So annuities aren’t just an interesting monetization method; they’re also a great way to keep retention up.

Annuity Tradeoffs: Cannibalization vs. Conversion

There are two important questions raised by annuities:

The first question revolves around the problem of cannibalization of sales. The other addresses the question of patience. Let’s take them in order.

Avoiding Cannibalization through Pricing Structure

Eternal Arena minimizes cannibalization through its pricing structure, which you can see below in Table 1. There is nothing else available besides the annuity in the $5 to $10 budget range. For people willing to spend $30 or more, annuities appear less like a substitute and more like a valuable rider on top of a lump sum purchase.

Table 1: Pricing of diamonds in NetEase's Eternal Arena

Of course, this does mean that NetEase misses out on people who would prefer to buy a lump sum in the $10 to $20 range, but here they are betting on the retention effect of the annuity to keep these players going and buying more over time. There’s also nothing preventing NetEase from offering lump sums in this price range after a player purchases the annuity.

Increasing Conversion by Segmenting Patient and Impatient Players

Annuities are also an effective way of segmenting players into two groups: patient and impatient. The longer you have to wait for something, the less valuable it is to you. This is a standard, intuitive assumption in economic analysis, which says that the present value of a future payment decays the further off it is in the future. The rate of this decay is called a discount rate. People with a relatively higher discount rate could be called impatient (and thus more likely to prefer buying a lump sum over an annuity) while those with a lower discount rate could be called patient (and thus more likely to prefer the annuity).

The Economics Behind Patient and Impatient Players

In the past, economists have assumed that the discount rate was constant over time, but experiments in psychology and behavioral economics have shown this assumption to be erroneous. In fact, we discount future payments at a non-constant rate. For example, many people will prefer to receive $50 now to $100 a year from now --- but they will often prefer getting $100 six years from now, instead of $50 five years from now. [5]

This reversal implies that we are very impatient in the near term, but the longer we have to wait, the more patient we become. We also tend to discount smaller payments at a much higher rate than larger payments, so we’re more willing to wait 15 days for $20 than we are willing to wait the same amount of time for $10 [3]. This is known as hyperbolic discounting, and it’s illustrated in Figure 2 below alongside the standard, constant discount. The literature on hyperbolic discounting finds that although we recognize the long-term value of being patient, we tend not to act that way in the short-term. Hyperbolic discounting has been used to model and explain procrastination, insufficient retirement saving, addiction, and overeating, among other phenomena. [4]

Table 2. In the same way that most people prefer $50 now versus $100 one year from now (due to constant versus hyperbolic discounting effects), Eternal Arena's annuity has less perceived value in the short term, but gains more perceived value over time.

So what does hyperbolic discounting tell us about the Eternal Arena annuities? First, since the daily payments are so small relative to what players can buy in lump sum – or what they need to make character/equipment draws in the game – we should expect annuities to effectively segment players by appealing more to patient types than to impatient types. This leads to higher conversion rates. Impatient players would be unlikely to go for the annuities option as a substitute for buying lump sums, while patient players would be unlikely to buy diamonds in the first place, if it weren’t for the annuity.

Second, all players – patient or impatient – will see longer-term value in the annuity. Again, this comes down to the discounting model. A constant decay model causes future payments to quickly drop to zero in present value, but the hyperbolic model slows this decay as the payments move further off into the future. This implies that impatient players will see annuity payments as of little value in the near term, but of great value in the medium to long term. So we should expect impatient players will add the annuity on as a rider in addition to – and not instead of – their initial lump sum payment.

A Promising Revenue Model in Search of More Games

In sum, annuities are a great alternative monetization method worth trying out. They effectively segment players into patient and impatient types, which increases conversion: The patient would not convert but for the availability of the annuity, and the impatient would convert whether or not annuities were offered. An effective pricing structure, such as that used in Eternal Arena, emphasizes the complementarity rather than the substitutability of lump sums and annuities. This minimizes cannibalization.

It certainly seems to be working for NetEase: According to App Annie, the Eternal Arena has consistently remained relatively high in iOS’ top grossing game charts for the past three months. With so many other games targeting a similar audience, I expect and hope we’ll see many more of them experiment with annuities in the future. Speaking of which, over on his own Gamasutra blog, my colleague Bill Grosso has more IAP strategies worth considering.

NOTES:

[1] http://mobiledevmemo.com/mobile-game-retention-rates/

[2] http://www.gdcvault.com/play/1014679/Revenue-Recognition-Game-Design-and

[3] Kirby, K. N. (1997). Bidding on the future: evidence against normative discounting of delayed rewards. Journal of Experimental Psychology: General, 126(1), 54.

[4] Frederick, S., Loewenstein, G., & O'donoghue, T. (2002). Time discounting and time preference: A critical review. Journal of economic literature, 40(2), 351-401.

[5] Thaler, R. (1981). Some empirical evidence on dynamic inconsistency.Economics Letters, 8(3), 201-207.