When SayGames announced its partnership with Korean developer NLABSOFT in November 2025, the two companies were already working on two projects together. One was the already-released 'Tiny Warriors Rush,' and the other, 'Zombie Fortress,' was in the testing phase. Both games were successfully launched to the global market last December.
First unveiled in December 2025, 'Zombie Fortress' quickly built a user base, recording over 10,000 reviews and an average rating of 4.3 stars.
With live services and further development currently underway, we take a look behind the scenes at the NLABSOFT development team, their development process, and how their partnership with SayGames is supporting the game's continued growth.
Zombie Fortress
From Core Vision to Finished Game
For NLABSOFT, 'Zombie Fortress' began with a very clear idea of the feeling the game should convey. NLABSOFT CEO Joo Jae-hyun stated, "What we are most proud of is that 'Zombie Fortress' perfectly realized the core idea of 'Zombie Crusher,' which was the name of our initial prototype." Joo added, "From the start, our goal was to create a game that is easy to play casually on mobile while maximizing the intense and satisfying thrill of wiping out swarms of incoming zombies."
While the zombie theme itself was familiar to global users, the team wanted to go beyond the traditional tower defense experience. By structuring gameplay around 'traps' instead of standard towers, they designed the core gameplay to focus on crushing, grinding, and blowing up incoming zombie hordes.
One of the most important parts of the development process was finding the right balance between 'intuitiveness' and 'depth.' In other words, they wanted to make the game intuitively understandable while ensuring that each session remained engaging as the complexity increased. "We invested a lot of time in balancing simplicity and impact," Joo added. "We are very satisfied with the current result."
Beyond simple game mechanics, the team hoped for Zombie Fortress to carry emotional weight. "We were inspired by 'The Last of Us' and chose a young girl as the protagonist. By placing her in a ruined world, we wanted to add a subtle narrative emotion to the game," Joo explains intuitively.
Turning Data into Discussions for Actual Product Development
Zombie Fortress went through numerous iterations during development, including two major pivots that fundamentally restructured the game. Initial prototype tests showed strong potential with positive retention and session lengths, but also revealed clear limitations in terms of onboarding for a broad global audience and the complexity of meta-content. Emphasizing this, Joo recalls, "At the time, it was difficult to clearly set the right direction with only limited internal data."
According to Joo, the collaboration with the publisher played a key role in systematizing product-related decisions. "Working with SayGames allowed us to approach improvements in stages," he noted. "We gained access to more extensive data, and by holding regular data-driven discussions, we were able to validate hypotheses instead of relying on guesswork."
SayGames producer Danila Katalev highlights the shared mindset behind this process. "I was very happy to meet a partner who shares a data-driven approach. We value the opportunity to review every test, every release, and every change to the product in detail together."
In the early version, users earned gold at the end of each wave, which they used to buy towers or upgrade archers to increase earnings per battle. While it functioned fine, this system created friction in the gameplay flow and made the onboarding process heavier than intended.
In the first pivot, they restructured the purchasing flow. Towers were moved to a bottom menu, and users were allowed to upgrade the probability of obtaining higher-quality tower options. While intuitiveness improved, it still relied on resource management.
In the second pivot, the game loop was completely simplified. Gold and purchase menus were removed entirely. Instead, after each wave, users were given the choice between a tower or a perk to strengthen their build. This reduced cognitive load, made the progression much more intuitive, and consequently helped significantly in strengthening key metrics.
These pivots were part of an intentional process to find growth points within the core loop. Each iteration led to clear improvements in funnel performance, retention, and overall lifetime engagement. By refining onboarding and simplifying progression, the development team strengthened key metrics and gained confidence for the global launch.
As the overall quality of the project improved and evolved, data analysis became a core tool for making product-related decisions. By February 2026, over 100 new analytical events had been implemented, greatly expanding the team's ability to observe, test, and refine the user's journey at every stage.
Danila explains this process: "The game is supported by 116 basic dashboards in our internal analysis system. In addition to this, we built 12 custom dashboards specifically for Zombie Fortress to better understand how users interact with the game. Our main goal was to provide as much data as possible to design and validate hypotheses for product improvement."
Through extensive A/B testing covering core features, monetization models, and content balance, the team was able to test hypotheses quickly and pivot with confidence. Since the global launch, the fact that one of the level balance dashboards has been accessed over 1,186 times clearly demonstrates the team's ongoing efforts to fine-tune the player experience.
Danila emphasizes that data analysis only plays its role when combined with a clear creative direction. "No matter how vast the knowledge base and advanced analytical technology you have, it cannot lead to success without a clear vision of what the game should look like," he says. "From day one, it was clear that the NLABSOFT team understood exactly what experience they wanted to deliver and what story they wanted to tell."
Joo also added, "SayGames provided objective feedback from the perspective of the global market while consistently respecting our creative vision."
The collaboration between the two companies went beyond just sharing reports or dashboards. Joo highlights the active, hands-on involvement of the publishing team. "The SayGames team, including management, played the game a lot themselves and provided very practical feedback on design and balance. Thanks to this, we gained the confidence to boldly stop, rethink, and redesign when necessary."
Danila shares similar sentiments. "The NLABSOFT team is more than just experts. Our product meetings often felt like a private fan club meeting that you could only join after investing thousands of hours in the tower defense genre."
Joo summarizes this dynamic partnership with a clear metaphor. "We built a sturdy ship and prepared a map to cross the Pacific Ocean toward a treasure island. SayGames helped us interpret that map correctly and served as our radar and lighthouse throughout the voyage."
Why Execution Speed is NLABSOFT’s ‘Superpower’
This iterative approach is closely aligned with the NLABSOFT team's way of working. Although the studio is small, it boasts a wealth of experience, consisting of members who have been developing mobile games together since 2010. As a result of working in sync for years, a tight-knit and highly reliable collaborative structure has been formed where teamwork naturally flourishes.
"Each team member is an expert in their field while also having the ability to handle various roles. Planners, artists, and developers participate from the early stages and make decisions together based on 'problem-solving' rather than job titles," Joo explains.
This flexibility shapes the studio's creative process. Planners don't just think about system logic; they constantly consider how a specific scene will be intuitively read even in a short marketing video. This mindset has a major impact on how the entire team evaluates and develops ideas.
Rather than refining concepts in isolation, NLABSOFT focuses on rapid implementation and validation. Regarding this, Joo explains, "We prefer to visualize ideas early and make decisions based on actual user feedback. Holding onto a concept for too long without testing only slows down development speed."
One approach that exemplifies this philosophy is the studio's 'Art Camp' method. Similar to songwriting camps in the music industry, this is a workshop where internal and external concept artists gather to explore various visual themes in a short, intensive manner.
The team sketches contrasting concepts such as 'Cavemen vs. Dinosaurs,' 'Farmers vs. Plants,' and 'Humans vs. Zombies,' and conducts CPI (Cost Per Install) tests for each. Joo explains, "By verifying which worldview and visuals resonate most with users, we decide the final art direction based on data rather than relying solely on intuition," emphasizing that "both Tiny Warriors Rush and Zombie Fortress were born through this process."
When asked about the strengths that define the team, Joo did not point to any single area. "If we had to define our superpower, it wouldn't be individual skill. It is the 'execution speed' driven by teamwork. In other words, it is the ability to quickly turn ideas into powerful prototypes and refine them into products that can be launched in the market without losing focus," he says.
For NLABSOFT, Zombie Fortress is not a point of completion, but a project that continues to evolve on the foundation of iterative production, trust, and a fast-moving team culture. As live services continue and the project evolves, the experience gained through the collaboration with SayGames is becoming a solid foundation not only for the development of the game itself but also for future projects they will create together.
Comments (0)