Since its launch in 2012, Candy Crush Saga has grown into one of the most successful tile-matching games in history, surpassing $20 billion in lifetime revenue by September 2023.
The evolution of King, from its acquisition by Activision Blizzard to joining Microsoft, has brought new opportunities for knowledge sharing and support. However, keeping Candy Crush as a forever game brings challenges, such as sustaining a massive player base and managing a game with over 21,000 levels.
At this year’s GDC Festival of Gaming, GamesIndustry.biz spoke with King’s head of live operations Eva Ryott to discuss this shift in player behaviour and how it sets the stage for King’s evolving strategy.
Ryott notes that players have less time to play games, and King is competing for that time. Players also spend more time on non-game activities, especially on mobile devices.
“What that means for us is we need to improve relevance for our games, we need to grab players’ attention immediately, and we need to ensure that we have an experience that captures the player,” she says.​
“Games are more and more complex these days. Candy Crush Live has been around for 14 years, which means we have more than 21,000 levels, lots of events and features. We’re looking at how technology can help us orchestrate that experience to make it relevant, to ensure that each player sees the right thing at the right time, but also how technology can iterate on the experience because the rate of innovation has gone up a lot in the market.”
Which is where AI comes in. Ryott clarifies that King uses AI to support workflows, but emphasizes that level designers use the technology as a tool to focus on creativity.
“We use AI and automation in our level and content pipeline, because managing over 20,000 levels – it’s not scalable to do that manually. Level designers use AI to identify levels that might break, be too difficult or too easy. It can also assist in the tweaking before we push the level out live to players. That means we put out higher quality levels sooner to the players, and then manual and mundane tasks are automated so our designers can spend more time on creativity and innovation.”
Ryott says investments in technology have become more important to King to support this initiative. “We used to be very quantitative data-driven, but because the market is more competitive, we need to dive deeper into understanding player needs and stay true to our audience, build trust, and truly understand their needs.”
She adds: “The other part is how King has evolved in that it used to be very much data-driven. But in the past five years, we’ve realised that in order to stay relevant and interesting to our players, we need to combine that with innovation as well and do larger expansion.”
Ryott highlights Candy Crush’s in-game tournament All Stars as an example of retaining players and attracting new ones, where players qualify for a live, in-person final to win a cash prize.
But how does King avoid overwhelming players with the features and content in a game that has run for over a decade?
Ryott says this isn’t an issue, as players aren’t bombarded with all 21,000 levels at once or the many features and events running daily. Players only see the section of the game they’re in and their progress. She notes that players are often waiting for new levels rather than sticking with the ones already provided.
“Last year we went from 45 levels per week to 60 levels per week because we had so many players waiting,” she says. “The way we then manage that scale is through AI and automation, which has enabled us to actually be able to drop more content to those players. So that’s where the technology really comes in to help us manage that.”
“When we launched Candy Crush, we could never have imagined it could be a forever game. Over time, we realised that, wow, we have a hit here. These players have been around since launch, and they’re still around. And so we have learned to build over the long term. And now we definitely look at it as a forever game.”
King’s main challenge is retaining and understanding its diverse player base.
“Understanding their different needs is very challenging, because they have different experiences and the way they play the game,” says Ryott.
“We want to keep the game familiar for these players because they like to stay in control when they play. They’re mastering the game, and they love strategising in the levels. But at the same time, we need to keep the game updated, fresh, and interesting for them.”
Ryott notes that longtime players have “certain expectations” of King, that Candy Crush is “bite-sized entertainment and that will always be the core.”
“That core gameplay is really important. The matching candies, that’s what players like and are coming to the game for. But then things around that, like events and features, that’s where we do the innovation. It’s really a balance that we always have in mind.”
Ryott uses the two-by-two Fish Booster as an example – a mechanic introduced last February. “It really showed us that you might think that ‘here is a great innovation, players will be happy, boom.'” [But] they’re like, ‘What is this?’ We really need to think carefully when we do big innovations. It’s needed and necessary, but we need to really understand what players expect from us constantly.
“We underestimated how these players would react because it turns out they noticed everything. And the way we found out was that we went to Reddit and there were so many complaints because the fish wasn’t smart enough, it was swimming towards the wrong level objectives and so on. That taught us we have this very loyal player base that we truly need to honour and listen to, not just look at the data.”
Ryott says King uses various methods to monitor player discourse, including community sites like Reddit, but mainly relies on traditional methods.
“Traditionally, we are a very data-driven company. We have built a lot of our live service on data-driven iterations. But over time, we realised we can’t just do smaller optimisations, we need to innovate as well to stay relevant in the market – we need to add more player research. So we have player researchers in the team, we interview players, we have focus groups, and we have various ways to gather those insights.
“Within Xbox, there are some really interesting learnings from Mojang in particular”
“But in this case, it was the forum where we actually discovered, because you get instant feedback. You don’t have to set up interviews; we got it overnight. After a few hours, when something is pushed live, you can go directly, and that’s the beauty of Reddit. But you also need to realise those are the most vocal players, so it’s not a good representation of it.”
As for King’s role in the wider Xbox organisation, Ryott says “there is knowledge sharing to be had” between the companies.
“We’re still developing the relationships, but we’re obviously excited to be part of this bigger family,” she says. “There are some really interesting learnings from Mojang in particular around how they spend a lot of time polishing things before they push it out to players versus we push something out and iterate, so that’s one takeaway.”
“We can definitely contribute with a lot of estimated learnings from mobile as a platform, but also live services, because we have built a lot of expertise around live services too.”
Ryott concludes: “We’ve been adopted by the right family, so to speak. We have the same kind of long-term thinking around games and expos, which are very focused on player needs, loyalty, and longevity. So in that sense, I feel like it’s a really good match.”
