2024 was a tough year for the gaming industry. Falling investment, rising overheads and player expectations combined to make developing and delivering a game more complicated and expensive. Everyone involved in the industry felt the impact.
In many cases, that impact was negative. But as I said in a previous blog on the topic of cost-management in gaming, difficult periods, like 2024, can lead to change for the better. I’ve already started to see that change manifest in the conversations our team has been having with game studios around game infrastructure earlier in the development process. We’re also seeing the benefits of hybrid infrastructure increasingly discussed as gaming companies look to reduce the impact of game infrastructure costs on their bottom line.
‘Survive til ‘25’ has been a popular phrase throughout the past couple of years. For those that know me, I’m more of a glass half full guy so I’d like to think that with some of the positive changes I’m going to discuss in this blog we could consider adjusting the phrase to ‘Thrive in 25’ and go into the new year with a more optimistic outlook.
Here are my predictions for three important gaming trends for 2025 that have one central pillar running through them: live service games (LSGs).
I went to Live Service Gaming Summit in London recently. The first of its kind, the event’s introduction to the gaming event calendar goes to show how big the subject of LSGs has become.
There are a couple of different LSG models. Firstly, there is the monthly subscription model, for which World of Warcraft is best known. There are a few other MMOs that follow this model, but they are few and far between. The second is games such as Fortnite, which are free to play with no monthly subscription, and instead make their money by selling in-game art – skins, backgrounds, special effects. These add-ons don’t impact game play at all but are an option for players to enhance and personalize their experience of the game. The third model is pay-to-win, which are typically more mobile games but do include some console/PC games.
While each LSG model approaches it in a different way, all are designed from the outset to provide game studios with a continual revenue model.
Some of the biggest games in the world are LSGs: Fortnite, Apex Legends, League of Legends, World of Warcraft, Destiny 2 and my favourite, Call of Duty: Warzone. To give you an idea of their success, Fortnite has generated over $26 billion in revenue and World of Warcraft has made over $9.23 billion.
These games have been around for a while, so why am I including a greater focus on LSGs in my predicted trends for 2025?
To make and launch a game successfully these days is hard. As one of our customers, Tomi Toikka from Makea Games said when we chatted recently – “it’s very unsafe territory to launch a video game”. Not only is it expensive and investment is hard to come by, it’s hard to make your game stand out amongst the over five million video games in the market and then retain gamer’s interest.
Game studios are increasingly looking at how they can create games that retain players for years rather than weeks or months. And it’s not just the studios that have a vested interest, as I’ve recognised from speaking with some of our game server hosting customers.
James Ross, CTO at Nodecraft explains, “a lot of games will be created and after a while they fall off in terms of popularity. One way of improving that is having some kind of player owned servers which allow people to play together and improve the longevity of the game.”
The other option is designing your game as an LSG. But if game studios choose to take this route, the question then is: how do game studios come up with new downloadable content (DLC) and new content without tipping the cost scales too much? As mentioned above, one of the big themes of this year has been cost management and that won’t be going anywhere anytime soon.
Game infrastructure plays a big part in helping studios find and maintain the right cost balance. Any DLC or new content launch is going to see a spike in player numbers followed by a drop after a period of time. Not necessarily back to pre-DLC drop numbers but you should always expect to lose some players.
The challenge for game developers who have created and launched an LSG is the ongoing infrastructure cost of keeping players around. The more players, the more infrastructure and therefore infrastructure spend is required. The key is finding the right balance between continuing to generate a profit whilst retaining players.
It helps hugely for a game studio to have a good understanding of both the predicted numbers for any DLC or new content launch, as well as the drop off pattern. Essentially, what percentage of the game infrastructure that is being provisioned for a DLC is going to be needed after two or three months? Game studios and companies can then better design and deploy infrastructure that can be flexibly scaled up and down in line with their players. Without a clear understanding of scaling requirements its incredibly easy for game studios and companies to overspend on infrastructure by overprovisioning or provisioning a set up that is designed for a greater level of scaling than they actually require.
The key is good forward planning. Some studios understand that; others are starting to get it, but I believe that in 2025 we’re going to see far more of these infrastructure conversations around live service games and games in general happen further in advance of new content launches.
A catalyst of these more advanced conversations is the move away from purely operating in hyperscale cloud to hybrid environments and my second key trend for 2025.
Nearly every conversation we have about game infrastructure revolves around how to create a successful hybrid set up. For one big reason: cost.
Hyperscale cloud has been a popular infrastructure option within gaming for some time now. But it’s unsustainable from a cost perspective and gaming companies are looking for alternative options.
Statistics around cloud repatriation within gaming specifically are lacking, so this prediction is based purely on the many conversations myself and our gaming specialists are having with our customers and others in the industry. Although general statistics support our findings – a recent survey of IT leaders showed that 93% had been involved in a cloud repatriation project in the last three years.
Infrastructure should be around 10% of a game’s overall budget. In my experience hosting game servers, hyperscale cloud is now counting for around 20-30% of a game’s budget. It’s a problem and the issue has partly been one of misconceptions around cloud, education and/or priorities.
As Liam Brennan, Director of Games Technology and Services at TENCENT said, “most people don’t know how to cost optimize from an infrastructure perspective”. Which is why we have had far more conversations with gaming companies and studios this year on the topic of hyperscale cloud alternatives and hybrid infrastructure and why I believe it’s going to take center stage in 2025. They need help figuring out the best hybrid infrastructure for their needs.
Let’s circle back around to LSGs as an example. If the developer of an LSG is accustomed to operating in a hyperscale cloud, they’ll be used to not having to plan their gaming infrastructure thoroughly. If they then need or want to move to a hybrid environment, some education will be required.
There are physical limitations with bare metal infrastructure, whether on-prem, co-lo or IaaS simply because, unlike cloud, you can’t have thousands of servers sat plugged in unused. That availability is a large part of why hyperscale cloud is as expensive as it is, as it’s essentially a luxury that is only ever really needed during peak hours. Which is why, to find a better cost balance when creating LSGs, game studios need to be considering timings around infrastructure conversations if they’re looking at going down a hybrid route, as many seem to be doing so now.
Indie game studio Makea Games and game orchestration provider Gameye are two great examples of companies that have successfully introduced a hybrid model.
“Our managed service is a curated fleet of servers that scales up, scales down to the studios as like a cloud but you’re all running on a fleet of bare metal machines. So, you’re getting that flexibility, but you’re not paying through the nose for it the same as you would do if you were completely dependent on a cloud vendor,” explains Andrew Walker, Head of Business Development at Gameye. “We use bare metal. We use metal as a service. We use hyperscale or cloud instances as an emergency back up, but by no means is it any one particular vendor. We deploy where the players are.”
Like Gameye, Makea Games also uses a hybrid set up of bare metal for their baseline CCU and Microsoft Azure for burst capacity. Makea Games Founder and CTO Ville Valtiala explains, “We looked at AWS and Microsoft Azure because they are popular options in the indie space, but they were very expensive. I knew at some point someone would turn around and ask me how much we were spending on infrastructure, and I didn’t want to have that conversation if we had gone fully down the hyperscale route.”
The importance of hybrid solutions is also being seen in industry research, with Omdia’s 2024 Cloud Platforms for Games report including bare metal cloud providers as challengers to the three big hyperscalers, of which servers.com was featured, for the first time this year.
We only have to read reports such as these and look at some of our partners who are developing solutions specifically for a hybrid world, such as Nitrado with Game Fabric, Gameye, 4Players with ODIN and Hathora to predict the continuing dominance of the hybrid infrastructure conversation in 2025.
To continue the theme of this post, my final prediction for 2025 is around AI, specifically its importance in LSGs but also more generally in the creation of game content.
Whether you think it’s going to have a positive or negative impact on the industry – and opinions are certainly split – I predict that AI is going to play an increasing role in gaming in 2025. Specifically, supporting the ongoing need to find more efficient and cost-effective ways of doing things.
As fantasy and science fiction author Joanna Maciejewska said in an X post earlier this year, “I want AI to clean my dishes and wash my clothes so I can focus on art and writing, not AI writing and drawing for me so I can clean dishes and wash clothes”.
Generating content for LSGs is incredibly work intensive for developers and therefore expensive. Generative AI in gaming can help speed up that process. As a gamer and reader of Metro wrote in to say, “rather than having dozens of people sat for hundreds of hours creating each individual building, character, tree, animal etc. all they’d have to create is a selection of reference material and leave it to the AI at the direction of the creative developers to build the world.”
AI can also be used for testing – to automatically detect bugs and defects in games and rank them according to priority and their impact on the game. Plus, the predictive capabilities of AI means it can be used to analyze data to predict potential issues within a game, allowing developers to take proactive action. Game developers can also use AI for player behavior analysis, allowing for more adaptive content and improved engagement based on real-time feedback, which are at the heart of sustaining long-term interest in LSGs.
As Hilary Mason, co-founder and CEO at game technology studio Hidden Door said in her talk ‘The joys of AI’ at Gamescom this year, “game developers become more like the directors of a play, rather than the actors or the writers. It’s more about deciding the arc and setting up all the pieces and allowing the systems and the players to provide the context in the details of the experience”.
There’s concern that the cost management benefits of AI may take away some of the work of developers, so views from the industry are understandably mixed. But if we look at it through my ‘glass half full’ lens, it’s also an opportunity to divert the cost of that saved time and spend that money elsewhere, on creating additional content, getting the game out into market quicker, and seeing greater success at launch.
If you look at the line up of game and console launches planned for 2025, it’s already promising to be a big one. Grand Theft Auto 6, Fable, Dune: Awakening, Monster Hunt Wilds and ARK 2 are all set to release along with a successor to the Nintendo Switch.
EY has predicted the gaming industry to generate around $211 billion in revenue next year and analysts are feeling confident about the industry’s outlook.
Mat Piscatella, video games industry advisor at Circana said, “the transition to the new normal in the video game market has not been painless. However, the long-term outlook remains optimistic, with 2025 in particular carrying significant growth potential.”
If the new normal for the industry includes a greater understanding of the role that infrastructure can play in balancing the cost scales, more conversations about alternative infrastructure set ups to hyperscale cloud with hybrid environments and an openness to the cost management benefits of AI, I believe 2025 could be a significant year for gaming.
Jarrod Palmer is our gaming industry specialist. He knows the issues our customers face and how best
to help them. He's also a great Warzone player and is the UK office FIFA champion.
