Game Economics, Part 3: Free-to-Play Games

This article is about why f2p games work and how you can use this information to add systems to a game to boost performance.

Apr 30, 2021

This article is about why f2p games work and how you can use this information to add systems to a game to boost performance.

This article is also about what free-to-play (f2p) games are and how they’re measured. Even if you’ve been around the f2p block a few times, I’m going to surprise you with a few new lenses and metaphors you can add to your toolbox.

Here’s a preview of some of what I’ll cover:

Why retention is the most important metric for long-term health
An introduction to the Live Games Trinity— the operational model I urge live games to adopt: regular content updates, live events, and effective merchandising.
The metaphor of economic voltage in f2p and some of the techniques you can use to turn up the voltage and generate more revenue
The profound impact of the power law distribution with respect to engagement and spending, along with a reveal of real-life data from games I’ve run that have never been shared before

This is Part 3 of a series on game economies. In Part 1 of this series, I covered the past and present of game economics, from the evolution of the industry from arcade machines and into a present that delivers more than half of its revenue through live game services. In Part 2, I covered the new world of digital collectibles based on non-fungible tokens (NFTs) and how it will impact games.

From Ugly Stepchild to Dominant Force

A little more than a decade ago, many of us in the f2p game industry heard from a multitude of game developers who didn’t believe these were real games; the story went that games ought to be packaged into experiences without any real-money interactions during the course of play (presumably, many of these critics had never inserted a quarter into an arcade machine). While there’s certainly a role for story-driven and stand-alone experiences that are freed from pecuniary concerns during the game, recent history has proven that most people prefer this business model. Today, free-to-play is 78% of the revenue in the game industry and is forecast to become as high as 95% by 2025.

Why F2P Works

Early critics generally missed the following advantages of free-to-play:

Democratization: Anyone can pick up a free-to-play game and opt-in to supporting it once they know they like it, which means the games can reach a far larger audience — and even the players who don’t pay can frequently enjoy them and become valuable members of the community.
Games as a hobby: Because of the ongoing nature of a game, players are more likely to think of the game as more than a game — as a hobby — which means a willingness to participate in the community around the game and invest in customization and collecting.
Better alignment with a service business model: Players are rewarded with a regular program of entertainment and content. Game-makers are financially incentivized to make the content.
More efficient monetization of attention: Generally speaking, the more someone plays, the more they’ll spend (vs. the inverse scenario of players paying less and less for the time they’re entertained in a “premium” game). Revenue per player is potentially uncapped.

The Live Games Trinity

Rather than thinking of a live game as a product with occasional changes, it can be more helpful to think of it as a program — not unlike a TV show — with a regular program schedule.

Why does this work? You’re nurturing your community by giving them things to do, stories to share, and events built around cooperation and competition.

It’s a flywheel built around social interaction, engagement, and converting scarcity into game revenue that can be reinvested in more content (which drives more engagement, and so on).

I’ll return to how you can employ the Live Games Trinity to improve almost any f2p game — but first, let’s go into some of the financial and economic principles that make these games work.

Financial Metrics

I’m going to focus on the metrics most frequently used in the internalized operations of a game (I’ll return to customer acquisition and discovery expenses in a future article). Here goes:

Retention. The measure of how long someone remains a customer, this is typically referred to as “dX” retention, where X is the number of days since signing up; the value represents the percent of installs who were active at that number of days. In other words, if d30 retention is 10%, that means “10% of the people who installed were still active on the 30th day.”
Engagement. There are many ways of measuring engagement, but for simplicity I’ll focus on a couple of high-level metrics: average session length (how long someone plays for each time they play) and the average number of sessions per day. Multiplying these two numbers together will give you the amount of time that the average person plays each day.
Conversion rate. Percentage of players who become paying players.
Daily Active Users (DAU). Average number of people who play in a day.
Average Revenue Per Daily Active User (ARPDAU). This is the total revenue collected from players in a day, divided by the number of players in the game. ARPDAU varies substantially between types of games; some games generate less than $0.10, while others can do $1 or more. Most companies report ARPDAU based on gross revenue (i.e., the amount actually paid by the player) before deducting any platform expenses (which currently range from 15% to 30% on mobile). In games with events, there can be substantial variability in ARPDAU (e.g., 2–3X or more during events).
Lifetime Value (LTV). Average revenue an individual player generates.

Retention and Engagement Are Most Important

If your retention falls off early, then you’ll cease monetizing your audience entirely. Players rarely return after extended periods.

While you’re retaining them, engagement determines how much you’ll be able to monetize.

Optimizing revenue is important, but a premature focus on revenue metrics will produce disappointing results. Mathematically, this is called optimizing a local maxima. The “global maxima” is the goal for any game:

The reason for this is actually fairly simple: think of any business as a system of constraints. There’s always something limiting growth or opportunity.

In a live game, you simply can’t monetize people who are no longer there.

Time Is Money

Or rather: Time isn’t money, but attention is. Almost all games are businesses that convert attention into revenue. Even “premium” games, which charge a fixed fee independent of the time played, are more likely to boost their revenue through downloadable content (DLC) updates and future sequels if a player was heavily engaged the first time.

Here is data from a game I previously operated, showing the number of sessions per day among different tiers of customers:

Yes, the top 10% of paying customers were connecting close to 60 times per day.

Are paying customers simply connecting more, or are the most engaged customers more likely to pay? It’s true that people are more likely to stick and engage after paying (something psychologists call the endowment effect. However, my experiences operating games have shown me that high engagement is the main precondition for players to make significant investments in the game — not simply the result of spending.

Voltage Metaphor

In physics, voltage is the “pressure” that causes electrons to move through a conducting loop.

Games have voltage too. Instead of electrically conducting pieces, it happens over the gameplay loop. And the way you generate the voltage is through the Live Games Trinity.

As someone plays a game, there’s a frequency with which they’ll encounter monetization opportunities: buy a currency, upgrade a character, collect a new item, view a rewarded-video ad. A player’s willingness to participate in these monetization opportunities — and the magnitude of these purchases — is the economic voltage inside a game.

You can compare this concept to the economic concept of the velocity of money. I prefer voltage as a more holistic metaphor — it includes the pressures and incentives to transact during the course of an experience.

Like actual voltage, it can vary over time. Many games look like this:

In other words, people here:

Monetize early
Continue to monetize, at tapering levels, for however long as they are retained (at which point, further contributions to LTV become zero)

The ARPDAU of a game in any given day is essentially the sum of all of the economic voltage graphs for all players for that point in time.

When the player is gone from the game, there’s no more economic voltage for them. Their player journey is complete, and you won’t earn anything more from them unless you have another game to introduce them to.

The most sustainable games are those which look better than the above graph. They exhibit very long retention, along with producing more economic voltage (jumps in monetization) along the player lifetime. More on that in a bit.

Power Law Distributions and Averaging Fallacies

Many game developers talk about LTV and ARPDAU, and these are helpful for doing some back-of-the-envelope math for a given game. However, they’re just averages: they don’t tell the real story.

*Never cross a river based on “average” depth. (Source:* The Flaw of Averages by Sam L. Savage, Hoboken, NJ: John Wiley & Sons, 2009.)

The big failing of an ARPDAU focus is that it doesn’t describe the journey of individual players through your game.

Segmenting your audience into the deeply engaged and high-spending players with some form of time-series analysis can help.

Understand that f2p games tend to exhibit what’s called a power law distribution. Here’s an example of a power law distribution of how the number of players in a game matches against the total LTVs in a game:

There are several important ramifications of this distribution:

As noted above, an “average” doesn’t really tell the whole story (if it did, we’d see a “normal distribution,” aka a “bell curve” formation, around LTVs).
There’s no cap on the highest potential LTV — mathematically speaking, the “mean” can actually be infinite.
A small number of players in the long tail of the distribution can generate an outsize portion of the overall revenue.

Here is the actual spending data from a game I operated:

In the above example, the top 20% of players spent 75% of the money (keep in mind that most players spend $0 in a F2P game). Twenty-four percent of the revenue came from the top 1% of players.

Whale Hunting

Now, some people make the wrong conclusions about the power law distribution phenomenon. For example, they assume that the entire f2p business is about “whale hunting.” Calling customers whales is a fairly contemptuous term — one that I don’t actually use. I also avoid some of the short-sighted conclusions that come from this belief.

Please: Be kind to whales. Photo by Wesley Tingey.

Here are some better conclusions:

Players who don’t monetize are also part of the community of a game. Interacting with these players is a form of content. Lose these players, and you may lose the energy of the community and cause the players who spend money to quit as well.
Don’t ignore long-term game mechanics. Players can play for years. Building mechanics that involve players for the long haul are a big part of the game’s sustainability. Frequently, these are features that embrace the idea that games are communities. Rather than whale hunt, build an ecosystem where all players can thrive for a long time.
Don’t cap potential spending. The focus here is on having a deep catalog value (the potential value of all the purchases someone could make) rather than simply hunting. The depth of the catalog is a greater constraint for most games than the existence of customers willing to buy (and unlocking more value is what will raise LTV and, consequently, enable more user acquisition spending at scale — a subject for another article).

All About That Scarcity (no Commons)

What makes a free-to-play economy work is scarcity:

Scarcity refers to the basic economic problem, the gap between limited — that is, scarce — resources and theoretically limitless wants.

Scarcity is fun to play with. Knowing that you’ve gained rewards, experiences, and virtual possessions that are in short supply (whether hard to earn or expensive to acquire) is satisfying. And scarcity is much of what generates the voltage of the Live Games Trinity.

As noted above, scarcity is about the gap between supply and demand. In f2p games, this includes:

Items, rewards, and currencies that are highly desirable.
Participation in rare experiences.
Reaching remote locations and/or deep progress into the story.
Being recognized as a “top” player.
Membership in good teams, social groups, and guilds.
Customizing your presentation to other players in a special way.

All the Right Drops in All the Right Places

Virtual property in a F2P game falls into two broad categories:

Currencies, which fulfill a similar role as currencies in the real world — namely, they’re a store of value and a medium of exchange. In other words, currencies are traded in the game for access to content or virtual items. Currencies are acquired via a real-money transaction or are earned within the game.
Items, which is everything else. Although you may be able to sell an item, their usual purpose is to provide benefits to the player.

Virtual Items

Here are some of the attributes that help you describe game items:

Utility: Items may have a utilitarian purpose within the game and offer advantages to the owner. For example, an item could increase the damage of the player, allow access to a region, or enable a certain type of crafting. In other words, having the item means you’re going to win more (however winning is defined within the game).

These dragon eggs were highly sought-after in a game I operated: they had both substantial utility AND affinity.

Affinity: Independent of the item’s underlying utility, some items may simply be liked by players. It is possible for an item to have zero utility but high affinity (e.g., items that players wear for social or cosmetic reasons, such as customizing the look of their avatar). Items could also have both affinity and utility, meaning they offer some sort of game-mechanical advantage and players want them for additional reasons, such as an attachment to a favorite character or how cool it looks. Depending on the community and gameplay, affinity could have as much or even greater contribution to value than utility.

Scarcity: Beyond the items that are simply given to every single player, items will have some form of scarcity. This scarcity may be limited due to progression (i.e., reaching a certain part of the game to gain the item) or other limitations on supply. A few ways of limiting the supply include setting a “drop rate” for an item (such as in loot or in a gacha mechanic), limiting an item to real-money purchasers, making an item available for a limited time (sales or events), requiring players to craft an item, or limiting the maximum number of the item that can exist in the game.

Value: A function of supply and demand. This determines what a player is willing to exchange for an item (which could include either paying currency to purchase the item, spending time to obtain it, or indirectly spending currency to accelerate the time required). Supply is determined by the scarcity of the item within the game’s design. Demand is a function of utility and/or affinity. The following surface graph illustrates how value will tend to increase according to these factors:

The slope of the value curve varies between games, but is frequently exponential. Game designers carefully balance their catalog of items to avoid providing wildly overpowered items relative to others of similar scarcity in the game.

A few other properties are worth identifying, as they impact utility and scarcity in interesting ways:

Path to Acquisition. How is the item actually obtained? As noted above, the path could be part of the game that’s hard to get to (a hard-to-reach area, a boss that’s hard to defeat), or the item could be commonplace, such as a starter item. Some items may only be obtained from other players, which depends on the trading rules of the particular game; in some games, the identity of the item’s creator or history of owner (its provenance) may also be important to players.

Transferability. Many games have rules that determine whether an item can be transferred to another player. This is usually a meaningless property in a single-player game. But in a multiplayer game, the transferability of an item has a large impact on the game’s economy. Many games simply don’t allow transfers at all (most free-to-play games), which allows the designers to carefully control the supply of items to each player. Others might allow trades of certain low-rarity or player-crafted items. And in other items built on realistic trading economies (such as Eve Online, or NFT-based games), virtually everything may be transferable.

Cooldown. This refers to how frequently an item may be used. Powerful items might be significantly limited in how often they can be used. Higher frequency items for the same power would have increased utility.

Charges. This refers to the number of uses an item has. Some items might be rechargeable; this property can be combined with other properties, such as a cooldown. Generally speaking, more charges will mean greater utility.

Duration. Similar to charges, an item might have a limited duration after which it is disposed of (or needs to be renewed). This is another way of placing constraints on certain items. For example, a “protection shield” in a war game might protect a city from incoming attacks for a limited period of time.

Progression Requirements. Some systems may require that the player reach certain levels (or whatever progression system the game is organized around) before an item can be utilized; this is to avoid premature advantages that might make the game too easy or boring.

Boosting Economic Voltage with Live Events

Remember the Live Games Trinity that I introduced a while back? Let’s return to one of its three pillars: events.

Events result in players’ spending becoming heavily engaged — at or above the levels they spent when they were experiencing the game for the first time. This helps avoid the tapering-engagement effect observed in many games.

This can often translate into substantial revenue improvements. In fact, I’ve seen live events boost ARPDAU in a game by 2–3x:

Events run for a particular period of time, generally from a few hours to a few days.

Events work because of scarcity, which can manifest in a number of ways:

Unique narratives: By participating in the event, you’re part of a story that is outside of the normal course of play.
Scarce experiences: Interact with the game in a unique way, such as unique contexts, game features, or even activities like the live music events that have launched in Fortnite.
Special drops: Items that are desirable to players are easier to obtain during the event.
Boosted and nerfed items: Some items may have their power levels boosted — for example, maybe fire damage is doubled during the event and ice damage is halved.
Special constraints: Particular items or characters may be required (or prevented) from participating.

Effective Merchandising

Events alone will drive significant engagement. The next step is to capitalize on this engagement by creating more opportunities for players to monetize, which in turn will fund an ongoing program of new content and additional events.

Some of the techniques include:

Event currencies that reward event participation with highly desirable items
Items and content offered as part of new storylines, characters, maps, etc.
Pairing events with limited-time offers for items that play a special role in extending the length of an event
Limited-time offers for items granted as rewards during or adjacent to events (to establish a high value perception associated with the items)
Personalized offers for players based on their purchasing history, interests, stage of their player journey, etc. Using machine learning, it’s possible to tailor offers down to the individual level.
Shopping interfaces organized around alternative currencies in the game, such as guild currencies rewarded for participating in cooperative events, arena currencies for participating in PVP, or other feature-specific currencies. Games like AFK Arena and Star Trek Timelines make ample use of these techniques.
Opportunities for non-monetizing players to participate in the economy through alternative methods, such as rewarded video ads.
Systems like VIP Systems to family purchases with special rewards, or Battle Passes that allow players to opt in to receiving a special class of reward items while they play.

Currency Types and Managing Inflation

One last economic principle that I ought to include at this point is the mention of inflation. Part of effective merchandising is to have the right currencies and the right items in each game, partitioned and segregated in ways that incentivize engagement rather than devalue the economy.

Almost every f2p game is an inflationary economy; currencies and items are not truly limited or scarce. This is why you don’t see unlimited player-to-player trading in most successful f2p games: the game’s owner needs to act as the monopoly supplier, trade regulator, and central bank. They mint new currencies and items whenever needed for a player’s experience — and because these items can be minted in an unlimited quantity, it becomes necessary to constrain the exchanges between players. Otherwise, advanced players would give currencies and powerful items to lower-powered players, and players would dump currencies and items on their friends when they’re exiting the game. Those behaviors would lead to a continuous devaluation of the currencies in the game, a massive power-creep in the items required to engage players, and a lack of interest in engaging with the core progression systems.

What’s the solution? In f2p games, it is limiting the number of functions that each currency can have. If one currency becomes devalued, it doesn’t have as much impact on the others. For example, an earnable currency is often the most inflationary because it becomes challenging to create enough currency sinks to extract it from the virtual economy.

That’s why at a minimum, games tend to have a dual-currency system where the “soft” earnable currency has a distinct purpose (e.g., allows purchases of common or more mundane items) from the real-money or “hard” currency (for getting the stuff people really want).

A third type of currency, which I already mentioned above, is an “event currency.” In general, this is perceived as a “score” (victory points, participation points, etc.) and can be organized into ranking systems (to incentivize competitive players) as well as threshold rewards (to incentivize more progression-oriented, “achiever” players).

The advantage of an event currency is that it has nearly automatic scarcity, determined by the time-limited aspect of the event being run. And inflation is managed, because each event essentially resets the currency back to zero.

Typically, event currency unlocks unique rewards that are hard (or even impossible) to obtain outside the event.

Successful games often have 10+ currency types to incentive specific behaviors. Beyond those already mentioned in this section, this can include:

Social currencies that reward getting your friends to play
Guild currencies that incentivize players to participate in cooperative activities
Inventory-limit currencies that cap the potential for certain currencies to inflate beyond controllable levels
Energy currencies to limit how much someone can repeat actions in the game
Feature-specific currencies to segregate additional ways of playing the game into their own sub-economies
Salvage currencies for destroying items you don’t want
VIP currency for extra perks for participating in real-money purchases (i.e., gamifying the purchasing system)

It should also be noted that many things that are perceived as game mechanics (levels, scores, experience points, etc.) are really currencies.

Content Train Don’t Stop

The Live Games Trinity is a methodology for live operations (aka games as a service) that enables growth and sustainability. It starts with putting in place a painless way to author, test, and deploy new content (items, stories, etc.) to a game to keep the Content Train running on a regular schedule. With that in place, you can start scheduling events that drive competition and cooperation within focused time constraints, which in turn can generate an ongoing revenue engine when paired with effective merchandising.

Metavert Meditations

Discussion about this post