Evolution of Game Content Industrialization and miHoYo’s Transformation Path

The global game industry entered a new phase of content industrialization in the 2010s, driven by the explosive growth of mobile gaming, rising user expectations for update frequency, and the surging cost of 3A game development. According to Newzoo’s 2023 Global Games Market Report, the global game market reached $183 billion in revenue in 2023, with mobile games accounting for 52% of the total, a share that has grown by 18 percentage points since 2015[^1]. This structural shift from premium console/PC games to live-service mobile and cross-platform games has fundamentally broken the traditional single-project workshop production model that dominated the industry for three decades, forcing all major game studios to rethink their production workflows, resource allocation, and asset management systems. miHoYo, a Shanghai-based studio that grew from a 5-person university dormitory team in 2012 to a global content giant with over 6,000 employees and four live-service IPs by 2024, is one of the most typical and representative cases of game content industrialization transformation in the past decade. Its evolution path not only reflects the core demands of global game industrialization, but also provides a unique solution to the core contradiction of the industry: how to achieve standardized and scaled production while maintaining the creative vitality of IP content.

Evolution Stages of Global Game Content Industrialization

The evolution of global game content industrialization can be divided into three distinct stages, each with clear characteristics driven by technological progress and market demand changes. The core driving force of each stage is to resolve the mismatch between production capacity and market demand under the constraints of existing technical and organizational conditions.

> 📌 侧边栏：什么是游戏内容工业化？ > 游戏内容工业化（Game Content Industrialization）是指将游戏研发、运营、发行的全链路流程拆解为可复用、可标准化的独立模块，通过工具化、自动化、资产化的方式，实现内容的规模化、稳定化生产，同时降低对核心团队个人能力的依赖。其核心逻辑与传统制造业的流水线生产一致，但保留了创意环节的灵活性，避免内容同质化。核心衡量指标包括：资产复用率、单项目研发周期、内容产出频率、人均产出、质量缺陷率。

Stage 1: Workshop Era (1970-2005)

The workshop era is the earliest production model of the game industry, with its core characteristic being that the entire production process is centered on the core creative team, with almost no standardized processes or asset reuse mechanisms. During this period, game development was equivalent to the personal creation of a small number of core members, and the success or failure of a project was almost entirely dependent on the ability and creativity of the core team.

Typical cases of this era include id Software’s *DOOM* (1993), which was developed by a team of only 15 people in 8 months, with all code, art, and design completed by the core team, and the total code volume was only about 100,000 lines. Another case is Nintendo’s *Super Mario 64* (1996), developed by a core team of only 4 people, including Shigeru Miyamoto, who personally drew all the conceptual design and level design drafts. During this period, the average team size for game development was less than 20 people, the average R&D cycle was 1 to 5 years, and the asset reuse rate between projects was less than 10%. After the project was launched, the core team would either disintegrate or turn to the development of completely unrelated new projects, and there was almost no accumulation of reusable assets.

The limitation of this model is extremely obvious: it can only adapt to a market environment with low product supply and high user demand. With the expansion of the game market and the rise of multi-platform releases in the late 1990s, the workshop model could no longer meet the demand for large-scale content supply, prompting the industry to begin exploring standardized production models.

Stage 2: Standardization Era (2005-2015)

The standardization era is the first stage of game content industrialization, with its core characteristic being the establishment of unified engine standards, asset libraries, and production processes, achieving scale expansion of single-project production. The core goal of this stage is to reduce R&D costs and shorten R&D cycles by improving asset reuse rates, so as to adapt to the market demand for high-frequency releases of series products.

Typical cases of this era include Ubisoft’s *Assassin’s Creed* series. Starting with the first release in 2007, Ubisoft established a unified Anvil engine and a cross-project 3D asset library, including architectural models, NPC animation templates, task system frameworks, etc. The development cycle of *Assassin’s Creed 2* (2009) was shortened from 3 years to 1.5 years compared to the first generation, the team size was reduced from 400 to 200, and the asset reuse rate reached 65%, reducing the R&D cost by 40%[^4]. Another case is Activision’s *Call of Duty* series, which established a unified IW engine and annual release mechanism after 2007, with a team size of only about 300 people, a development cycle of 1 year, and an asset reuse rate of more than 70%, achieving an annual revenue of more than $1 billion as early as 2012.

During this period, the average team size for 3A game development increased from less than 50 people in the workshop era to 200-500 people, the average R&D cycle was shortened from 3-5 years to 1-3 years, and the asset reuse rate increased from less than 10% to 30%-70%. However, this model still focused on single-project production, with low production frequency (mostly 1-3 years per product), and cross-project asset reuse was still limited to series products of the same IP.

Stage 3: IP Universe Industrialization Era (2015-Present)

The IP universe industrialization era is the advanced stage of game content industrialization, with its core characteristic being cross-media, multi-project parallel, and high-frequency update production under the IP universe strategy. The core goal of this stage is to maximize the value of IP assets through full-link content output, adapting to the market demand for long-term operation of live-service games and cross-media content expansion.

Typical cases of this era include Disney’s Marvel Cinematic Universe (MCU), which achieved an average of 2-3 movie releases per year after 2008, plus supporting TV series, animation, games, and other content, with total annual content output 5 times that of the previous single-movie model. Another case is Riot Games’ *League of Legends* universe, which has expanded from a single PC game to include mobile games, TV series, animation, comics, novels and other forms of content after 2015, with a total annual content output value of more than $3 billion.

The core technical support of this stage is the emergence of new generation game engines (such as Unreal Engine 5, Unity DOTS) and AI-assisted production tools. For example, Unreal Engine 5’s Nanite technology can support the direct import of film-level 3D models into games, reducing the model production time by 70%; AI concept generation tools (such as Midjourney, Stable Diffusion) can reduce the concept design cycle from 1 week to 1 day; AI code assistance tools (such as GitHub Copilot) can improve the programming efficiency by 30%-50%.

The following table clearly compares the core characteristics of the three stages of global game content industrialization:

| 阶段 | 时间范围 | 典型代表 | 团队规模 | 研发周期 | 资产复用率 | 内容产出频率 | 核心特征 | |------|----------|----------|----------|----------|------------|--------------|----------| | 作坊时代 | 1970-2005 | id Software、任天堂EAD | 5-50人 | 1-5年 | <10% | 1-2年1款产品 | 依赖核心团队个人能力，无标准化流程，资产几乎不复用 | | 标准化时代 | 2005-2015 | 育碧、动视、CDPR | 100-500人 | 1-3年 | 30%-70% | 1-3年1款产品 | 建立标准化引擎、资产库和生产流程，单项目规模化生产 | | IP宇宙工业化时代 | 2015至今 | 迪士尼、miHoYo、Riot Games | 500-6000人 | 6个月-4年 | 50%-80% | 每6周1次大版本更新，多项目并行 | 跨媒体、多项目并行、高频更新、IP资产全链路复用 |

The global game industry’s transition to the IP universe industrialization era has brought two core challenges to all studios: on the one hand, the demand for multi-project parall