Back in 2013, I wrote briefly about Games as a Service on this blog, again last year with Destiny as an example1 and by now this business model is quite dominant everywhere. I also gave some presentations on this and generally ranted about this to anyone who didn’t immediately run away. Meanwhile, a new hip buzzword has arrived and that’s Games as a Platform (GaaP).

However, what people mean with these terms remains a bit fluffy. In this post, I attempt to explain and define these terms better by comparing them to the services and platforms that exist elsewhere and I don’t try to pin them down by in-game features or naively and purely by their monetization model.

Promotional picture of Fortnite
Look at me playing with all my PSN friends

There are many ways to define and differentiate between these business models, but for me the natural way is to look at the developer’s income source and the player’s value source2:

  • “Traditional” games are products. The main source of income comes from selling the base game and additional income is earned through DLCs3. The game can offer service elements but this is not the main source of player’s value. Most AAA games remain here even today. Indie games remain here for nostalgic reasons or because they rarely can afford the backend services4.
  • Games as a Service offerings can have a price of entry, but a substantial source of revenue are additional purchases, in form of subscription payments or other recurring spending. The value proposition for the player comes from service elements, like multiplayer. Most free-to-play and multiplayer focused games are services.
  • Games as a Platform are still rare and the biggest barrier for them is that most games already run on a platform (Steam, Playstation, Xbox, Apple’s App Store…) who really like their monopoly power. The key differentiator to GaaS is that the main source of income comes from transactions and revenue share. Often, this means that players can buy and sell UGC.

Games as a Service

It is easy to say that all free-to-play games are GaaS or that GaaS means microtransactions. This is a bit unfair. It is also setting the bar a little low if every game with multiplayer is a service based game.

Just adding microtransactions to a game does not make into a GaaS title, but many GaaS titles have microtransactions. Microtransactions seem currently to be the most viable option for free-to-play games as ad-supported models might not generate enough income (unless you’re Voodoo or Ketchup, but that’s an entirely different business model) and subscriptions remain a hard sell (unless you’re World of Warcraft). However, there is some success in “event” or “season” subscriptions in games like Fortnite (Battle Passes) and CS:GO (Operation Passes).

The transition to games as services focuses on games as experience and not as a consumable thing. Many games, especially action/adventure AAA games, are still predominantly products - like films or books. You cannot open a Harry Potter book and expect the story to continue or be different than the last time you read it. You might see the content in a new light and so on, but the content has a definitive start and a definitive end5.

One way to define services6 is that they are

  1. Intangible - The game might still come in a box but contents of that box do not define the game
  2. Perishable - There is no way to “replay” the content
  3. Inseparable - There is no game without a server
  4. Variable - No two experiences are the same

As software, most games are intangible to begin with but a service-based game is not defined purely by its bits and bytes on a hard drive. Most games also provide an experience and as such could be considered perishable, but the key thing is that while you can replay a level in an Assassin’s Creed game you can’t really replay a certain PUBG match.

The third point relates strongly to “always online” in the sense that the game service is rendered to you by the server, Dota 2 or R6: Siege doesn’t make much sense without a server - they won’t even launch. For games, it’s really this third point of inseparability that defines GaaS - in good and bad. Once the servers go down, we are in danger of losing a good chunk of gaming history7.

Variability can be of course achieved without the game being a service, for example in systematic or open-world games like Far Cry 5, but these experiences are still somewhat fully designed for the player. The variability of service based games is a challenge for the developer as it would be great to be able to guarantee at least certain level of enjoyment. This can be difficult if your experience depends on other players8.

Taking the Harry Potter example further, GaaS is an attempt to be able to offer the player the Harry Potter universe and not just one (or more9) story within it. This is also where I think games are unique compared to other forms of media: We can provide worlds, or

We can leverage the interactive medium of video games with long-term retention, and manage the customer relationship lifecycle while focusing on the lifetime value of this relationship

Games as a Platform

The new trend is for game developers to transition their service offerings to platforms. The business case is simple: to further remove the developer or publisher from content creation and “outsource” that to the community of players and just keep running the money-making engine10. Competitive multiplayer games were suitable as a GaaS offering because the content or experience (the matches) are created by the players by just playing the game. However, they might not as easily transforms into platforms. Not just anyone can be allowed to create levels, weapons, or characters and break the delicate balance of the game. A lot of the value of the game comes from the quality of experience which can be easily disturbed by opening the gates.

Promotional image of Roblox
This game blindsided a lot of people but it just goes to show how success in GaaS doesn't have to be instant

There are many games which allow players to create content, especially on Steam through Stream Workshop, but this content is rarely treated as first class citizens in the games. Some content might be so good that it gets added to the game proper and the content creator might get compensated but the platform holder makes the decisions.

There are many examples of platforms outside of games: Netflix, Spotify, YouTube and so on. With games, it gets muddier real fast.

The main difference is that in GaaP games the 1st party (game developer) does not have a monopoly on the content or experience creation and on capturing the value of that content. UGC or multiplayer are common in GaaS games but even though these are created by the players, they usually do not have a possibility to monetize from these11.

The other key is that the game has to be able to act as a platform, ie. it is possible for 3rd parties to build on that platform. World of Warcraft is clearly a GaaS offering, whereas with Minecraft things get a bit muddier and Roblox could be considered a GaaP game.

One way to think about GaaP is comparing it to Steam or Apple’s App Store. The main source of income is the revenue cut3. They provide the platform, others bring the content.

Netflix can also be considered a platform in the sense that any movie or TV show can potentially be on their service, but they choose what they have on their service. YouTube is a better example as their gatekeeper duty is similar to Steam and Apple: anything goes as long as the content adheres to certain rules. It’s worth noting that both YouTube and Twitch offer subscription plans. For platform, like for a service, it doesn’t matter if the there is an up-front cost, subscription cost or if the service is offered for free (ad-supported or otherwise) but the key differentiator is that the platform holder doesn’t have a monopoly on content creation and revenue is shared. Sure, there is content created by Netflix, Apple and YouTube on their own platforms because the margins are probably better on those but these compete on somewhat equal footing with other 3rd party content.

It is important to note that GaaP means more than just player to player marketplaces. Games where P2P trading is expected from all players might not be as successful as games with more “realistic” expectation where some or even just few players are content creators and most of the players are consumers. The value added and so the revenue cut can be significantly higher in the latter case than in the first12.

The growing importance of streaming and eSports plays an interesting role for GaaP gaming. Can these forms of content creation be considered to be part of the “platform” of games such as Fortnite, Rainbow Six Siege or Dota 2? These streams and competitions are added value to the players and the content creators and competitors can earn money, however these flows do not pass through the game platform holder but it can be argued that the existence of this content is valuable to the platform.

Due to 1st party limitations, it’s difficult to see a future of GaaP on consoles and mobiles anytime soon. Another aspect that could be seen important for GaaP games is that they are platform agnostic (see YouTube which runs on anything connected to the Internet) and this is very much against the interest of the device-level platform holders13.


It is tempting to just slap GaaS label on your multiplayer game or GaaP on your game with UGC, but this is a very narrow and meaningless way to do it. These models explains both the business model and value model of the game. Being a platform requires releasing control and allowing others to invade your walled garden - too strict control will smother the potential of your platform. Content creation is a way of self-expression and you only get people to follow your rules if you compensate them well enough. At some point you’re just outsourcing.

The real challenge remains that way too often GaaS manifests itself to the player in a bad way: DRM that prevents legitimate players from playing, “Always Online” and its server errors, monetization models that feel aggressive or unethical, games perpetually in “Early Access” or Beta hell that still are happy to take your money, and countless 1st party installers and portals.

I would argue that this is not due to malice but unpreparedness14. Many developers were just unprepared for all the additional challenges that come with the service model. With products, it was okay to release a patch now and then so that the product works. With service based games, the players expect customer service, not just customer support. This increase in customer touch-points or interactions can easily overwhelm if you expected to only interact with the player when they give you their money in exchange for the game and not every time they launch the game.

The real transition has already happened but for most parts it was invisible to the players - the business model of the sausage factory should not affect player’s enjoyment. We realized that the players want experiences, not bits of code. We also realized players are willing to pay for experiences, not bits of code.

  1. My hope of Destiny 2 delivering on day one were unfortunately proved false.

  2. Rationally, in market economy a trade only happens when both parties expect to get more out of the deal than what they put in.

  3. However, it can be argued that for Apple as a whole, the main source of income is the hardware. 2

  4. I don’t mean this in a bad way, building and maintaining servers isn’t easy and that’s why Apple, Facebook, Google, Amazon and everyone else is offering developers free backend services like leaderboards and multiplayer. Although it’s work to emphasize they are free only in cost.

  5. An exception are more systems driven games that while not services, offer near-infinite amount of playtime. These are still products, just like Lego bricks and sets are.

  6. Kotler and Keller, “Marketing Management”, 2007

  7. For posterity’s sake, it would awesome if before going offline, these games were updated with the possiblity of anyone running the server. This is very unlikely to happen to most of the mobile games, for example.

  8. This could also explain why toxic communities are today such a big problem, not because people behave more badly in the post-truth world but because communities are evermore important to the experience level of a single player.

  9. With the Season Pass.

  10. This is even more enticing to shareholders who really like stable, recurring income and not fearing if the next iteration of a publisher’s flagship title doesn’t sell.

  11. For multitude of reasons.

  12. An apt comparison are credit card operators’ transaction fees and Apple’s or Steam’s revenue cuts.

  13. Fortnite and its availability and cross-play offer an intriguing glance at the power relationships between different platforms.

  14. ie. Hanlon’s Razor.