Copyright eurogamer
In 2018, a year after its launch, Fortnite made Epic $5.48bn. That's a staggering amount of money, and the company has continued to make billions from the game each year since. Other publishers, developers, and investors saw that, and quickly mobilised. But launching and maintaining a live-service game is hard. Yes, the genre continues to dominate player time (mainly thanks to giants like Minecraft, League of Legends, GTA Online, Roblox, and - of course - Fortnite), but plenty of games have failed to even come close to the status of these well-established 'forever games'. Last year, Sony's now-infamous Concord lasted just two weeks before being shut down and its studio closed. Sony remains committed to "diverse and resilient" live-service games, but admitted its strategy is "not entirely going smoothly". Marathon remains indefinitely delayed, while Fairgames seems to be in trouble. And Sony isn't alone. Warner Bros. Games continues to work on live-service games, despite its Suicide Squad flop. Remedy was "unsatisfied" with sales of its Control spin-off FBC Firebreak. Mech shooter Steel Hunters from Wargaming lasted just three months in early access. And even Sega has admitted releasing a live-service game is the company's biggest challenge. These are a handful of stories from this year alone. So, how do you find success with a live-service game? Why do some games succeed where others fail? "That is the billion dollar question," says Mike Morhaime, former CEO of Blizzard. Morhaime worked on both World of Warcraft and Overwatch, before releasing a new live-service game, Wildgate, earlier this year via his new studio Dreamhaven. "It is very difficult. You can see that by just looking at the numbers, the market is very saturated. There is no shortage of very good choices out there, and the top five list has been in that position for a very long time." So, how do you answer that billion dollar question? Is it a matter of right place, right time? Do you need to sustain interest in a game for the long-term? How important is it, really, to listen to fan feedback? And what exactly does success look like? I spoke to a number of developers across the industry to understand the complexities of live-service development, and try to figure out if it is at all possible to crack this impossibly valuable formula. To make those billions of dollars, monetisation is crucial. The free-to-play model has become the norm for many live-service games, with additional microtransactions generating cash. As Candy Crush Soda Saga bosses note, though, you need to respect player time. And they should know: King is perhaps one of the biggest success stories in live-service gaming, with Candy Crush grossing over $20bn since the launch of the franchise. "Stay with us forever, and then we cannot ask too much," adds general manager Paula Ingvar. "We're not asking for too much time. We're not introducing any paywalls, or any pay-to-win mechanics. It's free-to-play, free-to-win. You just make that part of your routine and a natural part of your life that gives you a bit of joy - without being a squeeze, or [generating] burnout." "We want our players to make our games a part of their lives for a really long time," says VP head of creative Abigail Rindo. "So sustainability is such a huge part of that, and just making sure that we're giving that right experience to players and keeping them engaged, sustainably, is incredibly important." Other live-service games stick to a premium model that relies on players paying up-front, then offering free updates. With its forthcoming live-service shooter Exoborne, developer Sharkmob has opted for a premium release. "There are different challenges, different expectations," says the game's executive producer Brynley Gibson. "It's really tough doing a free-to-play game. You need many more players. You get that naturally, but then to sustain them and keep them it's a different set of challenges." But with a premium game, players have already made an upfront investment, making it easier to retain their interest. "It's a different agreement with the consumer. We're saying there's enough game here for your money, and we promise to support it in a certain way. Free-to-play is a different agreement." Sharkmob co-founder Martin Hultberg adds the premium model comes with an extra side benefit: it can be a deterrent to cheaters. "In every online game out there, cheating is a big problem, and if it's free-to-play, you can just create a new account, create a new account, create a new account, keep jumping in," he says. "But if you actually have a cost upfront, you get rid of a lot of the returning cheaters because they simply can't do it. And so it helps to provide a higher quality level service by paying. That's why it's called premium, you're getting a premium service." Indeed, cheating can be a major strike against the success of a live service game. "As soon as cheating becomes rampant," Hultberg adds, "or as soon as that toxic community starts hijacking your experience, then you see an immediate drop-off. Why would you want to play against someone who's cheating?" Free-to-play games have also popularised certain expectations or conventions - Fortnite, for instance, has popularised the use of a battle pass. Is it better, then, to stick with what players are already familiar with in order to establish a live-service game? "I think the rule of thumb is to keep the established convention, but at the same time offer [something unique]," says Beomjun Lee, executive producer of free-to-play live-service looter-shooter The First Descendant from Nexon. "There are not many looter-shooters in the Korean market…That's why we studied a lot of other existing live-service games that are doing quite well. We learned a lot from them and we wanted to make sure that we learned from their strong points." Of course, copying the homework of others can be troublesome when plagiarism is concerned - players found icons in The First Descendant they believed were copies of those found in Destiny 2, which Nexon later removed stating the game "has been developed with deep affection and respect for various looter shooters". Conventions are "really important" to premium games too, says Gibson, as there's a rhythm people expect from live-service games. "There are areas to innovate, and there are areas to do what's expected," he continues. "If we look at what's established, we can say, 'Look, this is somewhere where we don't need to innovate'. Or if we do, it's a small 'i' - it's our take on it. It could be what makes sense for our game. But then we put our effort and our mental willpower into where we really want to push things, which are the unique elements." Ultimately, monetisation is intrinsically linked to the game's business model, Morhaime explains. "Where your business model relies on its ongoing monetisation is where your development team and designers are going to be incentivised to spend most of their time focusing," he explains. "And you want to make sure that lines up with what is going to produce the best ongoing experience for players." It's risky, then, when a business model and developer creativity are mis-aligned. Morhaime suggests a simple example of outfits for characters: if these are generating the most income, that's where effort will be focused. "But that's not necessarily going to continue driving engagement in the game," he says. "I think also the internal structure of the company and the development team is important in terms of how these decisions are being made, and where does the voice of the player sit in all of that, and how loud is that voice?" Indeed, unlike standalone games, live-service games rely on a continuous community of players, and ongoing dialogue with the developers. But how much should developers listen to players? When is player feedback actually useful? I asked Destiny 2 principal designer Alan Blaine during his GDC talk on Bungie's game how to make sense of feedback while still maintaining a vision for the game. "I think there's a skill in not looking for what players are asking for, looking at why they're asking for it," he responded. "There's always an internal problem, and it is not players' jobs to come up with a solution for it, although very often you have to figure out what's actually causing it. If you just implement every single thing that players ask for, you won't have a very coherent or fun game to play. "When something comes up against the priorities of the game - because the entire game will have pillars, goals, stuff like that - you have to then dig deeper into what's actually causing [the feedback]," he continued. "I don't think there's an easy answer to how do we do that, but you definitely don't want to just pick the things that players say they want. You want to understand what the root cause of the problem is." For Gibson, player feedback on Exborne is "vital". "If we want people to keep playing this game, engaging with this game, playing with their friends, then we do have to listen," he says. "But of course, we have to be true to our vision as well, what we believe is the right way. So we have our vision for what we want to achieve with the game, and then we open that [to players]." Using data from public tests ahead of its release, the Exoborne team can determine how players are playing and how that might differ from what they're saying in surveys. "That's where it's important to work out the motivations," says Gibson. "[Players are] not game developers necessarily. Sometimes they can be very sophisticated in their responses, but often they don't really know what's wrong. So you have to then sift through the data…and then through that filter, or our knowledge of making games over many years, we then interpret 'what is the right solution for Exoborne?'" Lee believes clear communication with The First Descendent players is imperative, citing regular live streams with the community as a key way of being transparent and gauging the response in real time. "Even when large numbers of users are asking the same thing, we try to communicate with the users very frankly if that's not really fitting our direction or vision," he says. "We try to be very honest when communicating about user feedback." Even then, players can be divided between accepting the response or still pursuing change, and the team will adjust direction accordingly. Within a game's playerbase, a key consideration is choosing to listen to hardcore players or more casual players, and balancing updates between them. "It's super important to keep the most engaged players engaged with the game, because those are also your evangelists," says Morhaime. "They're the opinion makers - the people who are trying out the games first and then telling their friends what they're playing. I think the more casual players are looking to see what the more hardcore players are feeling about the games. And so if they've already moved on, then they may be getting different recommendations." However, it's also essential to maintain approachable onboarding for new players, to provide an exciting initial experience. "If you stop being able to bring new people into the game," says Morhaime, "that is a very difficult situation, because you're going to have churn. You need to be able to replenish that and continue to grow the overall population of the game to stay relevant." As a relatively new game, The First Descendent has a smaller audience of hardcore players the development team is focused on appeasing, but larger updates must still carefully align content to cover all users. As a live-service game's popularity grows, that spectrum of audiences can grow wider apart - it's why Lee plans to add content for hardcore users while also making it easier for new players to reach endgame content "so that in the end they can play together". An established game will have an even wider spectrum of players. Ingvar describes Candy Crush as "easy to pick up and start, but it's difficult to master". Due to the mobile game's long-term success, that's led to a whole range of different players. There are new players who deserve "a gold standard first time user experience"; there are returning players looking to ease back in; there are active players seeking new features and events. And lastly there are end-of-content players, "our most loyal and devoted players", who can play a week's worth of content in three days. "With today's technology, we could easily pour content on them, but they actually like being in that end-of-content state of mind and moderating the pace," says Ingvar. "It really just comes down to understanding each of these player groups and what it is that they want or need," surmises Rindo, "which is why we talk to our players so frequently… Making sure that we understand where the player is in our game, and what it is they want from our game in that space, is really, really important." That spectrum of players can impede on storytelling, too. An exciting story could draw in new players and provide the impetus for existing players to stay long-term. Go heavy on the lore, though, and a live-service game can feel impenetrable and alienate new players. Typically, a story will have a beginning, middle, and end. But if a live-service game is meant to keep players engaged for the long haul, how do you tell a satisfying, yet ongoing story? "It can't actually end," says Morhaime. "It's a similar problem to some of these long-running [TV] series or comic books. I think there's a lot of effort that is spent on building the backstory and the characters themselves. And then there are story elements that are part of the narrative that drive different episodes of that, but there's overarching themes and arcs that can extend beyond these different chapters that are still interesting." Releasing approachable chunks of story as episodic content allows for shorter story arcs, while leaving enough open-endedness for future updates. It's an approach Sharkmob intends to take with Exoborne, taking inspiration from long-running TV shows. "We want to keep telling the story that we start in the beginning," says Hultberg. "We don't have an obvious end to it within what we call Chapter One. We're building chapter-wise content, so we'll keep adding to the story in the live phase. That also helps us support and justify the new content that comes out - instead of it feeling random or feeling disjointed from the product, you can help explain its place in our world through the narrative that we roll out in the live fix." What's more, a live-service narrative can evolve along with player tastes and preferences, with Hultberg joking about which characters could be killed off in future. "We don't want to tell the story of the people who fans don't really care about," he says. "We want to keep exploring the stuff fans are engaging with. I want to kill off one of the characters. I like killing off characters! But if they resonate, then we have to keep them alive." Focusing on character-driven stories is a further way to generate that episodic content without necessarily needing a continuous story, especially if that plays to a game's strengths. That's certainly how games like Apex Legends continue to operate as new playable characters are added to the game. Lee is aiming for a similar approach with The First Descendent, which over time has focused more on characters in-line with player interest. As such, "it's very important to set convincing background stories for each Descendant, and that's how we can make characters more likable for players," he says. "There is a big universe…but we use each episode to only partially disclose the narrative. In the end, players seem to really remember each drama and the conflicts of individual characters." Those character stories don't even need to be in the game itself. It certainly worked for Overwatch, as Morhaime explains: "Overwatch was an interesting one because there's actually very little story, maybe no story. There's story that's alluded to in the game… All of that happened outside of the game, but it was actually still very effective in giving you some reason to have affinity for these characters." One of the biggest strengths of a live-service game is putting players into the action and allowing them to tell their own stories. This has certainly been a major factor in Helldivers 2's success - a game that, in director Johan Pilestadt's words, embraces chaos. "We're trying to create a breeding ground for emergent stories to happen that you and your friend group or randoms online can have as an anchoring point of that moment you played Helldivers," he said in his GDC talk on the game's development. This is part of the reason friendly fire was included in the game. "Friendly fire is there for social reasons, not for anything else," he said. "Screwing up is part of the fun." Another key component is the Galactic War, where individual contributions count towards a greater goal. "The Galactic War is a super simple mechanic from a technical perspective but it harnesses the need for a community to be part of something greater to deliver some amazing moments," said Pilestedt. "Good design," he concluded, "is not about creating pre-made experiences; it's about creating systems where players discover how everything interacts." Emergent storytelling and gameplay is set to be a big part of Exoborne, too. "Player stories are, in a way, more important for us than for me as a narrative director to tell a story," says Hultberg. This could include putting players in interesting situations to make decisions, or encounter the unexpected, or narrative twists; but it's also about systemic gameplay. That's why Exoborne features extreme weather and forces of nature to interrupt players during a match. Within a game's systems, players then provide a wild card factor. "When you have any game, but especially a systemic game, then the players become one of the components to that system," says Hultberg. "How they interact becomes a key factor that can help you create these situations. We have to think about it when we create the game, but then also we have to respect the outcome of that, because it can generate results that are unexpected." Whether through storytelling, gameplay updates, or new monetisation options, a live-service game has the capacity to evolve over time and be refreshed. But how frequent should those updates be? Really, there's no set answer to that as it depends on the type of game, not to mention the cost and resources required for each update. And as Morhaime points out: "It's not so simple as just to put the spreadsheet down and say, 'if we double the cadence, we're going to double the revenue and double the engagement'. It doesn't always work like that." What it comes down to is reacting quickly to the community. For Sharkmob's previous game Blood Hunt, Gibson explains, "time to player" was key: "how quickly can we go from realising something or listening to the community, there's a clear need for something, and getting that into people's hands? If you can't minimise that time, really you're dead." Hultberg adds: "The service is for the people using it, not for the people developing it. So there needs to be a match of expectations." As Lee explains, there's a delicate balance between adapting to the community while still introducing quality content. "We want to keep the momentum for existing players, but at the same time we want them happy and satisfied with the new content we introduce," he says. "In the beginning, I was looking at the number of concurrent players going down, and I wanted to just do a weekly update so that we can retain them," he continues. "Then I learned while running a live-service game, it's simply impossible to do updates every week while providing quality content. And there are so many competitors, there are so many other games, so it's kind of selfish of me to expect players to only play The First Descendent always." Sometimes giving players space between updates allows for high quality additions to lure them back. "In order to run a healthy live service, the economics also have to work." For King, updates to Candy Crush Saga have changed over time along with player habits. It's less about the frequency of updates and more about the amount of content that's offered each time. Ingvar details further: "We definitely came from a place where we were constrained. We just couldn't meet [players'] demand and their appetite for more content. So we improved our processes, our tech, our tooling, and we added more people to the team to meet those needs and expectations. I think we've tripled or even quadrupled over the course of the last 10 years how much content we give every week, or every other week." The emphasis has also shifted from extending the game's longevity for expert players, to offering more variety to all players. "We've shifted our efforts, I would say 90 percent, to the existing progression, and making sure there's a nice mix of challenge, difficulty, different game modes, variety, introducing new things," says Ingvar. "We could just introduce new things to the end of content. That's not what we want. We want to introduce it across the entire progression." It should be noted, after this interview, reports emerged of King utilising AI for level design, with those tools replacing existing staff. King has not responded to Eurogamer's request for comment. Another factor for King is Candy Crush Saga being a mobile game - does this mean the team is constrained by technology when it comes to updates, in order to serve the widest audience? Rindo admits the team is sometimes constrained by technology or device, but what she loves about mobile is "we're always where the player is, it makes [the game] so accessible". This is where the design challenge comes from when it comes to providing updates: with such a broad range of players, it requires the team to be "a lot more creative around our design choices." "I think without hesitation, we would move with the players should there be another platform in the future, or some other cool device that comes along, as we have in the past," acknowledges Ingvar, noting Candy Crush was originally a web game, before being ported to Facebook and then mobile. "Right now I think the mechanics, the progression, the accessibility, all of that is such a good fit with mobile." As content updates bring fluctuations in player numbers, it's important to calculate concurrent players and retention as a measure of success - alongside basic finances, of course. "That number is real time," says Morhaime, "showing the health of the game, the size of the community, whether more people are choosing to play this game than they did last week." A healthy game will bring back players. "A live service game needs to be generating ongoing revenue to be able to subsidise the ongoing cost of the live service," Morhaime surmises. "In order to run a healthy live service, the economics also have to work." Lee sees retention rates in The First Descendent as "the most important metric". With each update, the team will also analyse the percentage of returned users compared to the previous update. "It's most important because it shows how much the players trust the dev team's updates," he says. Gibson suggests player sentiment as a core metric, a game's "theme score", in addition to tracking how deeply players are interacting with specific systems in the game. For Hultberg, though, sustainability for developers is of primary importance. "It needs to be sustainable for the team, both in terms of revenue for a company, but also for the team's ability to output work and not grind themselves into a pulp while doing it," he says. That sentiment is mirrored by Rindo. "I want our teams to make work that they're proud of," she says, "because that's also hard in live-service games. You're working on the same game for so long, you want to make sure the team is engaged because if the team's making stuff they're proud of, then players get amazing things as well." Even with strong metrics, how can live-service games compete with those dominating giants? Is there a point where a live-service game naturally comes to an end? "Given the fact that the span is literally between one week and two decades for some games, I would think it depends on how the entire industry and genre evolves," says Hultberg. "Some games people just stop playing because they're no longer interested, and you can't really do anything about that." And while there's stiff competition within the games industry to launch a successful live-service game, there's competition from outside the industry too. As Pilestedt explained, the idea of games-as-a-service is making a game into a hobby, not just a standalone product. For him, that means competing with the likes of Warhammer and Magic the Gathering, while Hultberg says live-service games are competing for time with TV shows and social media too. "It used to be a couple of years ago that we were mainly focusing on competition from other games, and it was mostly on a wallet level - how much money do you have to spend this month?" he says. "For me right now, it's how many hours can you consume entertainment every day, and how do you fit your game within that three or four hour period? "That is a very tough competition, because we know a lot of people actually prefer to watch over playing [games]. It is more fun to watch on YouTube than to play so that's definitely a concern. I mean, interesting challenge." "Some games people just stop playing because they're no longer interested, and you can't really do anything about that." "Interesting challenge" is under-selling it. Between monetisation options, storytelling, and constantly providing meaningful updates to a highly critical audience of players, finding success with a live-service game is a near impossible task. Helldivers 2 is perhaps the biggest success in recent years, but as Pilestedt explained in his GDC talk, that wasn't down to price point, or emergent gameplay, the marketing, content frequency, or even launch timing. It was all of these things in tandem. No wonder his talk was titled "Capturing Lightning in a Bottle". So how do you do it? "The only way you're going to get there is you have to produce a very high quality experience," says Morhaime. "I don't think you can get there by being a clone of one of those games, because if your game is too close the existing population is going to choose the game they're already playing. There is a switching cost, and also their friends are there, and moving all of their friends is going to be really difficult." He continues: "There's got to be some new innovation. It used to be, we would look for some big problems that we thought we could solve, [but] these games are very good, so there's not a big obvious problem. You have to be looking at doing something unique and different and innovative and still a very high quality and polished experience." He admits that some low budget indie games have managed to offer unique experiences, and due to their low cost the bar is lower for success and profitability. "But if you're going to go and create a live-service game competing with some of these very successful, top games," he says, "I think you're going to have to spend some money." The bottom line is that live-service success isn't easy, despite the billions of dollars that could be gained. As these developers have proven, success relies on a complex web of business and creative decisions - and, perhaps even, a little spark of luck.
