数字平台生态系统跨类别创新战略及演化：技术驱动视角的案例研究
Cross-category innovative strategy and evolution of the digital platform ecosystem: A case study from the perspective of technology-driven

摘要：数字平台生态系统的战略与演化是数字化时代重要的管理议题。本研究聚焦跨类别创新这一关键战略行动，以字节跳动为典型案例，基于分类理论构建“技术驱动—战略行动—生态演化”的研究框架，解析以数字技术为核心驱动，通过战略设计与组织适配推动跨类别创新的阶段性演化过程。研究发现，在类别创建阶段，平台通过身份战略、合法性战略及独特性战略，形成以核心业务为基础的生态系统；在类别跨越阶段，平台借助平台包络战略、开放式创新战略及地位战略，构建跨类别协同的生态系统，实现技术能力溢出与场景融合创新。研究进一步分析了跨类别创新战略与生态系统的共演机制：技术驱动下的战略行动不仅重塑市场分类，更通过数字资源与新类别的整合推动生态系统的结构化升级，生态系统的动态演化又为跨类别创新提供了新的战略空间。研究创新性地提出跨类别创新战略框架，明确了关键战略行动与平台发展的耦合机制，为构建突破路径依赖、可持续发展的数字平台生态系统提供了有益参考。
Abstract: The strategy and evolution of the digital platform ecosystem are important management issues in the digital age. This study focuses on cross-category innovation as a key strategic action, taking ByteDance as a typical case, and constructs a research framework based on classification theory of "technology-driven—strategic action—ecological evolution" to analyze the phased evolution process of cross-category innovation driven by digital technology, through strategic design and organizational adaptation. The study finds that in the category creation stage, platforms form an ecosystem based on core businesses through identity strategy, legitimacy strategy, and uniqueness strategy; in the category crossing stage, platforms build a cross-category collaborative ecosystem with platform envelope strategy, open innovation strategy, and status strategy, achieving technology capability spillover and scenario-based innovative integration. The study further analyzes the co-evolution mechanism between cross-category innovative strategy and ecosystem: technology-driven strategic actions not only reshape market categories but also promote the structural upgrading of the ecosystem through the integration of digital resources and new categories, while the dynamic evolution of the ecosystem provides new strategic space for cross-category innovation. The study innovatively proposes a cross-category innovative strategy framework, clarifying the coupling mechanism between key strategic actions and platform development, providing valuable references for building digital platform ecosystems that break path dependence and achieve sustainable development.

关键词：数字平台生态系统；跨类别创新；平台战略；类别演化
Keywords: digital platform ecosystem; cross-category innovation; platform strategy; category evolution

1. 引言
1. Introduction

随着数字技术的快速发展，数字平台生态系统已成为驱动产业变革和经济发展的重要组织范式。数字平台企业作为核心主体，依托云计算、大数据、人工智能等前沿技术，构建起高交互性、强开放性的平台架构，互补企业依托数字平台，通过整合或调用互补资产实现协同创新，进而构建多元化的产品或服务，用户通过数据生成、消费行为及使用反馈等形式参与生态的建设与优化（Hein et al.，2020）。数字平台生态系统凭借网络效应快速扩大用户规模，沉淀数字资产并提升商业价值。传统平台的竞争优势面临着数字技术加速迭代的威胁，数字平台的通用性、即时互联性与可扩展性能够实现超越传统产业边界的合作与竞争。在此背景下，行业的颠覆者可能不是同行，很有可能来自完全不同的领域。例如，亚马逊、谷歌等在电商、云计算和人工智能等领域建立主导地位。平台若过度依赖特定竞争优势和成功路径则容易陷入“伊卡洛斯悖论”，即因固守传统优势而错失新的发展机遇（Miller et al.，1990）。如何突破路径依赖、实现生态系统持续演化，成为数字化时代重要的管理议题。
With the rapid development of digital technology, the digital platform ecosystem has become an important organizational paradigm driving industrial transformation and economic development. Digital platform enterprises, as core entities, rely on cutting-edge technologies such as cloud computing, big data, and artificial intelligence to build high-interactivity and highly open platform architectures. Complementary enterprises leverage digital platforms by integrating or accessing complementary assets to achieve collaborative innovation, thereby creating diversified products or services. Users participate in the construction and optimization of the ecosystem through data generation, consumption behavior, and usage feedback (Hein et al., 2020). The digital platform ecosystem rapidly expands user scale and accumulates digital assets, enhancing commercial value. The competitive advantages of traditional platforms face threats from the accelerated iteration of digital technology. The generality, real-time connectivity, and scalability of digital platforms enable collaboration and competition beyond traditional industry boundaries. In this context, the disruptor in the industry may not be a peer but could very well come from completely different fields. For example, companies like Amazon and Google have established dominance in areas such as e-commerce, cloud computing, and artificial intelligence. If a platform overly relies on specific competitive advantages and success paths, it may fall into the "Icarus Paradox," where clinging to traditional strengths leads to missing new development opportunities (Miller et al., 1990). How to break path dependence and achieve continuous evolution of the ecosystem has become a crucial management issue in the digital age.

跨类别创新作为突破传统行业分类边界的战略行动，通过整合不同类别的技术优势、商业模式与制度逻辑，能够创造出全新的市场机会并重塑竞争格局，有助于理解数字平台跨行业发展的新特征。分类理论提出，类别具有社会结构的基础框架作用，能够分析组织结构分区、内外互动关系以及重构变革过程（Navis & Glynn，2010）。但传统的产权结构及交易成本理论难以解释数字平台跨行业发展的特性，数字平台通过类别创建与类别跨越等方式突破组织产权和交易的边界，重构用户的类别认知（唐方成等，2023），创造打破固化行业结构，形成具有市场竞争力的新类别。这一理论为解析数字平台生态系统如何通过跨类别创新突破路径依赖，提供了从物质本体特征分析的独特视角（Kennedy & Fiss，2013）。
Cross-category innovation, as a strategic action that breaks the boundaries of traditional industry classifications, can create entirely new market opportunities and reshape competitive dynamics by integrating the technological strengths, business models, and institutional logics of different categories. Category theory suggests that categories serve as a foundational framework for social structure, enabling the analysis of organizational structural divisions, internal and external interaction relationships, and the process of reconstructing transformation (Navis & Glynn, 2010). However, traditional property rights structures and transaction cost theories struggle to explain the characteristics of digital platform cross-industry development. Digital platforms break the boundaries of organizational property rights and transactions through category creation and category crossing, reconstructing users' category perceptions (Tang Fangcheng et al., 2023), creating new categories that disrupt rigid industry structures and form competitive market forces. This theory provides a unique perspective for analyzing how digital platform ecosystems break path dependence through cross-category innovation (Kennedy & Fiss, 2013).

基于此，本研究以分类理论为基础，构建“技术驱动—战略行动—生态演化”的研究框架，试图解答以下核心问题：一是数字平台的跨类别创新战略包含哪些关键阶段？各阶段的核心战略行动如何体现技术驱动特征？二是探究数字平台通过哪些战略行动，跨越不同类型平台之间的壁垒重塑生态结构？三是跨类别创新战略与生态系统如何实现协同共演，实现生态的可持续发展。为了回答以上问题，本研究通过对中国互联网科技公司字节跳动展开纵向探索，选择字节跳动的依据是公司从算法推荐驱动的内容平台起步，逐步跨越至短视频社交、电商、云服务及通用人工智能等差异显著的领域，完整呈现了数字平台以技术为底层支撑，通过战略行动实现跨类别创新的发展过程。研究为构建突破路径依赖、跨行业可持续发展的数字平台生态系统提供了有益参考。
Based on this, this study constructs a research framework of "technology-driven—strategic action—ecological evolution" based on category theory, attempting to address the following core questions: First, what are the key stages of cross-category innovation strategies for digital platforms, and how do the core strategic actions in each stage reflect technology-driven characteristics? Second, how do digital platforms, through strategic actions, cross the barriers between different types of platforms to reshape the ecological structure? Third, how do cross-category innovation strategies and ecosystems achieve synergistic co-evolution to achieve sustainable development of the ecosystem? To answer these questions, this study conducts a longitudinal exploration of ByteDance, a Chinese internet technology company. The choice of ByteDance is based on the company's initial focus on algorithm-driven content platforms, which gradually expanded into significantly different fields such as short video social networking, e-commerce, cloud services, and general artificial intelligence, fully demonstrating the development process of digital platforms achieving cross-category innovation with technology as the underlying support through strategic actions. This study provides valuable insights for building digital platform ecosystems that break path dependence and achieve sustainable cross-industry development.

2. 文献回顾
2. Literature Review

2.1 分类理论下的跨类别创新
Cross-category innovation under the theory of classification

在组织理论中，分类理论回应了一个经典问题：专一领域的深耕还是跨类别的发展会得到更高绩效？Zuckerman（1999）的研究中率先提出，组织需符合行业类别的认知框架以获取合法性，否则可能被资本市场忽视，研究强调了组织适应特定类别的必要性。众多学者围绕该议题展开讨论，形成了对后续研究影响颇深的分类规则（Category imperative）。分类规则将类别视为外部制度环境的核心构成，并将类别与评论家、监管者及消费者等对不同类型组织施加的影响联系起来（Hsu & Hannan，2005），其作用机制体现为受众对组织的双模式评估过程，即组织传达了属于特定类别特征的文化“代码”，受众首先判断组织所属类别，然后确定其符合类别预期的程度（Phillips & Zuckerman，2001；Durand & Paolella，2013）。虽然符合观众的期望通常是明智的，但后续研究发现，最大的回报可能流向那些通过创建新类别和在相应界面进行创新的人（Zuckerman，2017）。Pontikes（2012）指出，受众可以高度接收混合、跨越和延伸类别的组织，甚至视其为创新的信号，有助于组织的创新发展。跨类别创新来自组织链接以前未合并的类别，通过整合不同的物质元素、商业模式组件或制度逻辑，形成挑战现有分类系统并创造新市场空间的战略行动（McDonald & Allen，2022）。Durand和Khaire（2017）根据跨类别是由现有分类系统之外的元素还是由内部元素形成，区分了跨类别形成的两种模式：类别涌现和类别创建。吴小节等（2020）根据分类的演化过程补充说明了类别跨越与类别再分类两种情境。跨类别创新可以及时调整组织的身份及形式以适应新兴市场的竞争压力，通过身份构建、合法性获取、独特性创新和地位提升等具体的战略行动塑造新的价值创造模式。跨类别创新标志着组织未来形态的改变，相关研究及应用有助于促进市场中的技术变革、产业演化及颠覆式创新等。
In organizational theory, the theory of classification addresses a classic question: does in-depth cultivation in specialized fields or cross-category development lead to higher performance? Zuckerman's (1999) research first proposed that organizations need to conform to the cognitive framework of industry categories to gain legitimacy, otherwise they may be ignored by the capital market, emphasizing the necessity of organizations adapting to specific categories. Numerous scholars have discussed this issue, forming a classification rule (Category imperative) that has a profound impact on subsequent research. The classification rule treats categories as the core composition of the external institutional environment and links categories to the influence imposed by critics, regulators, and consumers on different types of organizations (Hsu & Hannan, 2005), its mechanism is reflected in the dual-mode evaluation process of the audience towards the organization, that is, the organization conveys the cultural "code" of belonging to a specific category, and the audience first judges the category to which the organization belongs, and then determines the degree to which it meets the category expectations (Phillips & Zuckerman, 2001; Durand & Paolella, 2013). Although it is usually wise to meet the audience's expectations, subsequent research has found that the greatest rewards may flow to those who create new categories and innovate in corresponding interfaces (Zuckerman, 2017). Pontikes (2012) pointed out that audiences can highly accept organizations that mix, cross, and extend categories, even viewing them as signals of innovation, which helps the innovative development of organizations. Cross-category innovation comes from organizations linking categories that were previously not merged, by integrating different material elements, business model components, or institutional logics, to form strategic actions that challenge existing classification systems and create new market spaces (McDonald & Allen, 2022). Durand and Khaire (2017) distinguished two modes of cross-category formation based on whether the cross-category is formed by elements outside the existing classification system or by internal elements: category emergence and category creation. Wu Xiaojie et al. (2020) supplemented two situations of category crossing and category reclassification based on the evolutionary process of classification. Cross-category innovation can timely adjust the identity and form of the organization to adapt to the competitive pressure of emerging markets, shaping new value creation models through specific strategic actions such as identity construction, legitimacy acquisition, unique innovation, and status enhancement. Cross-category innovation marks the change in the future form of organizations, and related research and applications can promote technological transformation, industrial evolution, and disruptive innovation in the market.

2.2 数字技术驱动下的跨类别创新
Cross-category innovation driven by digital technology

数字平台生态系统作为平台数字化的演进形态，其本质是通过数字技术构建价值共创网络。该系统的技术架构设计范式受数字技术的特性驱动。其中，模块化特征使系统具备动态重构能力，分布式架构则支持系统规模的跨边界拓展，数据智能分析确保信息交互的标准化和可复用。
The digital platform ecosystem, as an evolutionary form of platform digitalization, essentially constructs a value co-creation network through digital technology. The technical architecture design paradigm of this system is driven by the characteristics of digital technology. Among them, the modular feature enables the system to have dynamic reconstruction capabilities, the distributed architecture supports the cross-boundary expansion of the system scale, and data intelligence analysis ensures the standardization and reusability of information interaction.

（1）数字技术驱动生态边界拓展
(1) Digital technology drives the expansion of ecological boundaries

数字技术通过模块化和分布式架构为数字平台生态系统跨类别创新提供底层资源支撑。物理技术层面，平台通过模块化设计提升产品开发效率，依托分布式架构构建内部可重复使用的核心基础，支持第三方软硬件接入（Gawer & Cusumano, 2014），如SAP云平台通过开放应用程序开发接口连接物流、财务等垂直领域服务商，形成企业应用生态系统。这种架构为跨类别创新提供了两大优势：一是通过技术复用降低准入成本。平台无需为新业务重新开发底层技术，通过调用已有模块即可快速搭建新服务。二是基于核心业务发展跨类别创新业务。平台可在不影响核心业务的前提下，通过微服务化改造试点新功能，逐步培育用户认知后再整合至主界面，如微信通过小程序拓展电商、本地生活服务等，或是通过与标准化技术协议促进与互补企业间的创新协同，发掘更具创新性的组件和技术（Kwak et al.，2018）。上述方式催生了平台包络战略的数字技术实现路径，当平台通过技术复用积累跨类别能力后，可通过功能捆绑或并购等途径实现对相邻市场的快速渗透（Eisenmann et al., 2011），形成从数字技术能力溢出向生态边界拓展的正向循环。
Digital technology provides underlying resource support for cross-category innovation in digital platform ecosystems through modularization and distributed architecture. At the physical technology level, platforms enhance product development efficiency through modular design and build reusable core foundations based on distributed architecture, supporting third-party software and hardware access (Gawer & Cusumano, 2014). For example, the SAP Cloud Platform connects logistics, finance, and other vertical service providers through open application development interfaces, forming an enterprise application ecosystem. This architecture offers two advantages for cross-category innovation: first, it reduces entry costs through technology reuse. Platforms do not need to redevelop underlying technology for new businesses; instead, they can quickly set up new services by calling existing modules. Second, it enables cross-category innovation based on core business development. Platforms can pilot new features through microservice transformation without affecting core business operations, gradually cultivating user awareness before integrating them into the main interface, such as WeChat expanding e-commerce and local life services through mini-programs, or fostering innovative collaboration with complementary enterprises through standardized technology protocols to discover more innovative components and technologies (Kwak et al., 2018). The above methods have given rise to the digital technology implementation path of the platform envelope strategy. When platforms accumulate cross-category capabilities through technology reuse, they can rapidly penetrate adjacent markets through means such as feature bundling or mergers and acquisitions (Eisenmann et al., 2011), forming a positive cycle from digital technology capability spillover to ecological boundary expansion.

（2）数据驱动用户认知重构
(2) Data-driven reconstruction of user cognition

用户对数字平台跨类别创新的接受程度与技术和认知的耦合水平密切相关。首先，数据智能分析能够充分复用数据并降低认知复杂度，计算社会科学将定量大数据算法分析与定性叙事发展相结合，能够从实时交互数据中识别行为模式差异，将跨类别的产品和服务转化为符合用户既有认知的形态，推动用户接收跨类别创新（Cudennec & Durand，2023），并以此为基础优化新平台的服务方案和匹配效率。其次，社交传播能够触发用户认知迁移，利用用户生成内容、社交裂变以及共创叙事重塑类别在用户中的语义共识，实现类别的认知跨越（Glynn & Navis，2013）。三是基于技术背书的用户认知合法性构建，跨类别创新需解决用户对新服务身份的信任质疑，通过技术认证与权威协同可赋予认知合法性（Ocasio et al., 2015）。例如，天猫国际通过区块链技术实现商品溯源，将生态边界延伸至跨境电商可信服务，增强了用户对新类别的信任度。
User acceptance of cross-category innovation in digital platforms is closely related to the coupling level of technology and cognition. First, data intelligence analysis can fully reuse data and reduce cognitive complexity. Computational social science combines quantitative big data algorithm analysis with qualitative narrative development, enabling the identification of behavioral pattern differences from real-time interaction data, transforming cross-category products and services into forms that align with users' existing cognition, thereby promoting user acceptance of cross-category innovation (Cudennec & Durand, 2023). This serves as a foundation for optimizing new platform service solutions and matching efficiency. Second, social communication can trigger cognitive migration, utilizing user-generated content, social viral marketing, and co-creative narratives to reshape the semantic consensus of categories among users, achieving cognitive leaps across categories (Glynn & Navis, 2013). Third, the construction of cognitive legitimacy based on technological backing is crucial for cross-category innovation, as it needs to address user trust in the identity of new services. Technological certification and authoritative collaboration can grant cognitive legitimacy (Ocasio et al., 2015). For example, Tmall Global extends its ecological boundary to cross-border e-commerce trusted services by using blockchain technology for product traceability, enhancing user trust in new categories.

2.3 数字平台生态系统跨类别创新
Cross-category innovation of the digital platform ecosystem

现阶段有关于数字平台生态系统跨类别创新的研究较少，相关研究集中在创新生态系统跨界的讨论上。基于组织生态理论与资源依赖理论的整合视角，创新生态系统的跨界行为可界定为：在位企业依托既有生态系统领域的生态位优势，通过构建动态搜索机制识别潜在创新合作伙伴，进而形成跨边界资源整合能力（Walrave et al.，2018）。这种组织间协同创新行为本质上是企业通过重构价值网络关系，突破传统资源约束边界的创新实践，如企业借助生态系统网络成员间的复杂交互关系，实现异质性创新资源的跨领域整合。数字平台生态系统跨类别创新是一种战略行动，与跨界创新的差别在于从类别的视角映射平台、用户及评价者之间的关系，通过引入不同类别的技术、商业模式及制度逻辑等要素，解决价值创新过程中的资源整合与认知建构等问题。跨类别创新不仅需要平台与互补企业间的业务协作，还需要考虑数字化时代，参与者类型的指数级增长、交互模式的智能化演进、价值创造的非线性叠加以及治理机制的去中心化转型（Benitez et al.，2023；Pershina et al.，2019）。数字平台生态系统划分为交易型生态系统、信息型生态系统和创新型生态系统等多种形式（Cennamo，2021）。交易型生态系统的核心功能在于构建数字化连接系统，通过硬件基础设施层实现供需双方的精准匹配；信息型生态系统作为信息传播基础设施，其价值创造机制体现为通过算法推荐、语义分析等技术实现信息的智能分类与动态分发；创新型生态系统的核心特征表现为构建通用技术架构体系，通过模块化接口标准、开放API协议及协同开发机制，实现核心产品与互补创新模块的深度集成。三种类别的数字平台在功能定位、协同方式及价值创造方面的差别很大，在这三者之间跨类别创新可能带来新功能和新商业模式（Mann et al.，2022），避免过度依赖特定竞争优势和成功路径，但同时增加了管理不同类型平台的复杂度。因此，实现跨类别创新需要通过技术、数据、组织与制度等多方位共同驱动生态位重构，平衡组织层级化管理与开放的分类化演变之间的战略稳定，以技术能力沉淀为基础，保持创新战略行动能力，推进生态系统螺旋式上升发展。
Currently, there is relatively little research on cross-category innovation in the digital platform ecosystem, with related studies primarily focused on discussions about the boundaries of innovation ecosystems. From an integrated perspective of organizational ecology theory and resource dependence theory, the cross-boundary behavior of innovation ecosystems can be defined as: incumbent enterprises leveraging the niche advantages of their existing ecosystem domains to construct dynamic search mechanisms to identify potential innovation partners, thereby forming cross-boundary resource integration capabilities (Walrave et al., 2018). This interorganizational collaborative innovation behavior is essentially an innovative practice where enterprises reconstruct value network relationships to break through traditional resource constraint boundaries, such as enterprises achieving cross-domain integration of heterogeneous innovation resources through the complex interactive relationships among members of the ecosystem network. Cross-category innovation in the digital platform ecosystem is a strategic action, differing from cross-boundary innovation in that it maps the relationships among platforms, users, and evaluators from a category perspective. By introducing elements such as different categories of technology, business models, and institutional logic, it addresses issues of resource integration and cognitive construction in the process of value innovation. Cross-category innovation not only requires business collaboration between platforms and complementary enterprises but also considers the exponential growth of participant types, the intelligent evolution of interaction models, the nonlinear accumulation of value creation, and the decentralized transformation of governance mechanisms in the digital era (Benitez et al., 2023; Pershina et al., 2019). The digital platform ecosystem can be divided into various forms such as transactional ecosystems, informational ecosystems, and innovative ecosystems (Cennamo, 2021). The core function of transactional ecosystems lies in constructing a digital connection system, achieving precise matching between supply and demand through the hardware infrastructure layer. Informational ecosystems, as information dissemination infrastructure, create value through mechanisms such as algorithmic recommendations and semantic analysis to achieve intelligent classification and dynamic distribution of information. The core characteristic of innovative ecosystems is the construction of a general technology architecture system, achieving deep integration of core products and complementary innovation modules through modular interface standards, open API protocols, and collaborative development mechanisms. The three categories of digital platforms differ significantly in their functional positioning, collaborative methods, and value creation. Cross-category innovation among these three types may bring new functions and business models (Mann et al., 2022), avoiding over-reliance on specific competitive advantages and success paths, but it also increases the complexity of managing different types of platforms. Therefore, achieving cross-category innovation requires multi-faceted common driving of niche reconstruction through technology, data, organization, and institutions, balancing strategic stability between hierarchical management and open categorized evolution, maintaining innovative strategic action capabilities based on technical capability accumulation, and promoting the spiral upward development of the ecosystem.

综上所述，数字平台生态系统与跨类别创新的研究逐渐成为学术焦点，尽管相关研究已为本文提供了一定的概念与理论基础，但目前的综合研究较为匮乏。数字平台生态系统具有超越传统产业边界发展的能力，但缺乏关于如何运用数字技术驱动及跨类别创新战略应对相应的情境变化和竞争压力。跨类别创新涉及多个领域技术、制度和资源的整合，是突破传统创新边界的重要途径，但现有研究对其内在机制的探索不够深入，无法为数字平台的创新实践提供有力的理论指导。因此，本文以字节跳动为案例，探究数字平台生态系统跨类别创新战略选择与演化过程，并总结出相应的理论模型，为数字经济背景下的平台创新战略和分类理论研究做出贡献。
Overall, the study of digital platform ecosystems and cross-category innovation has gradually become an academic focus. Although related research has provided this paper with certain concepts and theoretical foundations, comprehensive studies are currently lacking. Digital platform ecosystems have the ability to develop beyond traditional industry boundaries, but there is a lack of understanding on how to use digital technology to drive and respond to corresponding situational changes and competitive pressures through cross-category innovation strategies. Cross-category innovation involves the integration of technology, systems, and resources from multiple fields, serving as an important way to break through traditional innovation boundaries. However, existing research has not explored its internal mechanisms in depth, failing to provide strong theoretical guidance for innovation practices in digital platforms. Therefore, this paper takes ByteDance as a case study to explore the selection and evolution process of cross-category innovation strategies in digital platform ecosystems, and summarizes a corresponding theoretical model, contributing to the research on platform innovation strategies and category theory in the context of the digital economy.

2.4 研究框架
2.4 Research Framework

本研究基于分类理论，构建 “技术驱动—战略行动—生态演化”的研究框架，如图1所示，聚焦数字平台生态系统跨类别创新的阶段性战略及演化机制，解析技术驱动下战略行动与生态系统的共演逻辑。研究将数字平台跨类别创新划分为类别创建与类别跨越两个阶段，各阶段通过特定战略行动推动生态系统从核心业务奠基向跨类别协同升级。在类别创建阶段，聚焦技术优势构建与用户认知合法性塑造。数字平台依托算法推荐、用户画像等核心技术的模块化封装，通过系列跨类别创新战略，构建以核心业务为基础的信息型生态系统。此阶段的关键是通过技术创新突破既有分类系统，形成新价值类别的市场认知。在类别跨越阶段，借助技术—业务双轮驱动突破类别边界。平台通过通过系列跨类别创新战略，实现技术能力跨场景复用与多源数据整合。此阶段的核心是通过技术架构与战略设计的耦合，推动信息型生态系统向交易型、创新型平台演化，实现技术溢出与场景融合创新。平台的价值创造模式从平台技术与价值的独立涌现转向多元主体共创转变。该框架基于分类理论揭示数字平台如何通过技术模块化、战略阶段性设计与生态协同，突破路径依赖并构建可持续竞争优势，为解析数字平台跨行业发展的内在逻辑提供了理论指导，也为企业分阶段实施跨类别创新战略提供了实践指引。
This study is based on category theory to construct a "technology-driven—strategic action—ecological evolution" research framework, as shown in Figure 1. It focuses on the phased strategies and evolutionary mechanisms of cross-category innovation in digital platform ecosystems, analyzing the co-evolution logic between technology-driven strategic actions and the ecosystem. The research divides digital platform cross-category innovation into two stages: category creation and category crossing. Each stage promotes the ecosystem's upgrade from core business foundation to cross-category collaboration through specific strategic actions. In the category creation stage, the focus is on building technological advantages and shaping user perception legitimacy. Digital platforms rely on modular encapsulation of core technologies such as algorithmic recommendation and user profiling, through a series of cross-category innovation strategies, to construct an information-based ecosystem based on core businesses. The key at this stage is to break through existing category systems through technological innovation, forming market recognition of new value categories. In the category crossing stage, it breaks through category boundaries with dual driving forces of technology and business. The platform achieves cross-scenario reuse of technological capabilities and integration of multi-source data through a series of cross-category innovation strategies. The core at this stage is to drive the evolution of the information-based ecosystem towards transactional and innovative platforms through the coupling of technology architecture and strategic design, achieving technology spillover and scenario integration innovation. The platform's value creation model shifts from the independent emergence of platform technology and value to co-creation among multiple stakeholders. This framework, based on category theory, reveals how digital platforms can break through path dependence and build sustainable competitive advantages through technology modularization, phased strategic design, and ecological collaboration, providing theoretical guidance for analyzing the internal logic of digital platform cross-industry development, and also offering practical guidance for enterprises to implement cross-category innovation strategies in stages.

图1 研究框架
Figure 1 Research Framework

3. 研究设计
3. Research Design

3.1 研究方法
3.1 Research Methods

案例研究作为探索复杂管理现象的重要方法，能够通过情境分析描述、测试和生成理论，并深度探讨理论与实践的互动机制（Yin, 2018；Eisenhardt，1989）。案例以其实践性与承载性的属性优势，现已广泛运用于管理科学、社会科学等应用性学科（苏敬勤等，2021）。鉴于数字平台生态系统跨类别创新的动态性与情境依赖性，本研究拟采用探索性单案例研究方法，方法选择的核心依据包括：（1）理论探索需求：对跨类别创新的技术驱动机制及战略行动过程缺乏系统性解析，单案例研究能够聚焦特定现象，深度解构关键构念与作用路径；（2）现象典型性：单案例研究有助于总结和归纳跨类别创新的动态过程，分析多业务板块的跨越式发展为观察技术驱动下的战略行动提供了独特窗口，能够有效地呈现不同阶段的变化特点；（3）过程追踪优势：单案例研究可以针对典型企业纵向数据收集，深入捕捉技术驱动与战略决策的内在关系，解析技术驱动下跨类别创新战略与生态系统的共演逻辑。据此，本研究遵循探索性单案例研究的步骤和规范。
Case study, as an important method for exploring complex management phenomena, can describe, test, and generate theory through contextual analysis, and deeply explore the interaction mechanism between theory and practice (Yin, 2018; Eisenhardt, 1989). With the advantages of practicality and bearing capacity, case studies have now been widely used in applied disciplines such as management science and social sciences (Su Jingqin et al., 2021). Given the dynamism and context dependency of cross-category innovation in digital platform ecosystems, this study proposes to adopt an exploratory single-case study method. The core basis for method selection includes: (1) theoretical exploration needs: there is a lack of systematic analysis of the technical driving mechanisms and strategic action processes of cross-category innovation, and single-case study can focus on specific phenomena, deeply deconstructing key constructs and pathways; (2) typicality of phenomena: single-case study helps to summarize and generalize the dynamic process of cross-category innovation, analyzing the leapfrog development of multiple business segments as a unique window for observing strategic actions driven by technology, which can effectively present the characteristics of changes at different stages; (3) process tracking advantages: single-case study can collect longitudinal data from typical enterprises, deeply capture the intrinsic relationship between technology driving and strategic decisions, and analyze the co-evolution logic of cross-category innovation strategy and ecosystems under the driving of technology. Accordingly, this study follows the steps and norms of exploratory single-case study.

3.2 案例选择
3.2 Case Selection

案例选择字节跳动的主要原因在于以下三个方面：一是企业发展过程与研究理论的适配性。字节跳动作为以算法推荐为核心技术起点的数字平台，从信息分发场景迁移至短视频内容推荐，进一步衍生出个性化商品推荐的电商平台，本质上是技术能力外溢与场景重构的过程。数据智能分析能够深入了解用户认知偏好，通过数字技术复用和数据复用将跨类别创新转化为可感知的、符合用户认知的产品，则能够有效地迁移数字能力至新类别市场，提升用户新体验。二是企业为技术驱动型跨类别创新实践的代表性。字节跳动集成了数字平台的大部分功能，是少数实现了跨类别生态系统全域布局的企业之一，其业务覆盖信息型（今日头条、抖音）、交易型（抖音电商）、创新型（火山引擎、豆包）三类生态系统，且各类型平台均形成显著市场影响力，完整地呈现了数字平台通过技术赋能实现跨类别创新的演进路径，具有显著的行业代表性。三是多源数据来源支撑研究的深度追踪。字节跳动作为国内知名数字平台企业，受到研究机构和媒体长期跟踪采访，能够获取的资料包括公司年报、产品白皮书、行业研究报告及媒体深度报道，完整呈现企业技术布局与战略决策脉络，丰富的信息为本文的数据收集与验证提供了保障。
The main reasons for selecting ByteDance as a case study are threefold: first, the adaptability of the company's development process to research theories. ByteDance, as a digital platform with algorithmic recommendation as its core technology, has migrated from information distribution scenarios to short video content recommendation, and further derived an e-commerce platform for personalized product recommendations. This is essentially a process of technological capability spillover and scenario reconstruction. Data intelligence analysis can deeply understand user cognitive preferences, and through the reuse of digital technology and data, it can transform cross-category innovation into perceivable products that align with user cognition, effectively transferring digital capabilities to new category markets and enhancing user experience. Second, ByteDance is representative of technology-driven cross-category innovation practices. ByteDance integrates most functions of a digital platform and is one of the few companies that have achieved full-scale layout across cross-category ecosystems. Its business covers three types of ecosystems: informational (Toutiao, Douyin), transactional (Douyin E-commerce), and innovative (Huoshan Engine, Doubao), and each type of platform has formed significant market influence, completely presenting the evolutionary path of digital platforms achieving cross-category innovation through technology empowerment, with significant industry representation. Third, multi-source data sources support in-depth tracking of the research. As a well-known domestic digital platform company, ByteDance has been tracked and interviewed by research institutions and media for a long time. The materials that can be obtained include company annual reports, product white papers, industry research reports, and media in-depth reports, completely presenting the technological layout and strategic decision-making context of the company, and providing rich information for the data collection and verification of this paper.

3.3 数据收集
3.3 Data Collection

本研究数据收集过程遵循“三角验证”的要求，涵盖一手访谈资料、二手资料和参与式观察等。具体如表1所示。以二手资料为先导，主要涵盖公司的年报、招股说明书、行业研究报告、文献和媒体资料等方面，建立对案例企业的初步了解。在此基础上，利用研究团队与案例主体的紧密联系，对企业开展了访谈调研。考虑到本研究属技术与战略议题，因此，采用半结构化方式访谈了技术、管理团队和部分用户，对字节跳动技术和管理业务展开业界与学界的对话式研讨。访谈分为3个部分：第一，询问工作经历、业务发展历程以及参与者对研究主题的认识；第二，详细阐述战略选择的考虑过程，包括技术研发决策、跨业务协同机制及生态发展策略等，同时询问了替代方案舍弃的原因，梳理共性因素、结构和思考框架，与访谈者进行理解对焦；第三，考虑到数据收集的充分性和完整性，对产品功能模块进行使用，同时调研部分用户的使用反馈，以保证观察的稳健性。以上3个方面资料形成交叉验证提升研究信效度。
The data collection process of this study follows the requirements of "triangulation," covering primary interview materials, secondary materials, and participatory observation, among others. Specifically, as shown in Table 1. Secondary materials are used as a starting point, mainly including the company's annual reports, prospectuses, industry research reports, literature, and media materials, to establish a preliminary understanding of the case company. On this basis, leveraging the close connection between the research team and the case subject, interviews were conducted on the company. Considering that this research belongs to the field of technology and strategy, semi-structured interviews were conducted with technical, management teams, and some users to carry out industry and academic dialogues on ByteDance's technology and management business. The interviews were divided into three parts: First, asking about work experience, business development history, and participants' understanding of the research topic; second, elaborating on the consideration process of strategic choices, including technology R&D decisions, cross-business collaboration mechanisms, and ecological development strategies, while asking about the reasons for discarding alternative solutions, sorting out common factors, structures, and thinking frameworks, and conducting understanding alignment with interviewees; third, considering the sufficiency and completeness of data collection, product functional modules were used, and the feedback of some users was investigated, to ensure the robustness of observation. The above three aspects of materials form cross-validation to enhance research reliability and validity.

表1 数据收集与编码概况
Table 1 Data Collection and Coding Overview

3.3 数据分析
Data Analysis

立足于探索性纵向案例研究，本研究的数据分析范围自公司2012年成立至2024年的核心业务演化，重点关注与技术驱动直接相关的跨类别创新事件，分为类别涌现和类别跨越两个阶段。在第一阶段，整理收集到的数据，按时间线划分阶段进行历时性编码，并构建“历史事件库”，系统梳理字节跳动的发展历程，并识别了各关键事件的时间节点，如图2所示。
Based on exploratory longitudinal case studies, this research's data analysis scope covers the evolution of the company's core business from its establishment in 2012 to 2024, with a focus on cross-category innovation events directly related to technology drivers, divided into two stages: category emergence and category crossover. In the first stage, the collected data were organized, staged according to the timeline for diachronic coding, and a "historical event database" was constructed to systematically review the development history of ByteDance, identifying key event time nodes, as shown in Figure 2.

图2 字节跳动发展历程
Figure 2 ByteDance Development History

在第二阶段，遵循Gioia等（2013）提出的归纳式质性研究方法指导数据分析，采用一阶、二阶的数据结构进行概念化编码，通过分析归纳聚合形成理论式主题。首先从数据中识别出与字节跳动跨类别创新相关的叙述，通过结构化编码方式对叙述进行归纳总结，完成从原始数据到一阶概念、二阶主题再到聚合维度的层层抽象。最终构建了完整的字节跳动跨类别创新战略及数字平台生态系统的演化过程，如图3所示。
In the second stage, guided by the inductive qualitative research method proposed by Gioia et al. (2013) for data analysis, first-order and second-order data structures were used for conceptual coding. Through analysis and induction, theoretical themes were aggregated. First, narratives related to ByteDance's cross-category innovation were identified from the data. Using structured coding methods, the narratives were summarized, completing the layer-by-layer abstraction from raw data to first-order concepts, second-order themes, and aggregated dimensions. Finally, the complete evolution process of ByteDance's cross-category innovation strategy and digital platform ecosystem was constructed, as shown in Figure 3.

图3 数据分析结构图
Figure 3 Data Analysis Structure Diagram

4.案例分析与发现
4. Case Analysis and Discovery

4.1 类别涌现阶段：技术赋能核心业务
4.1 Category Emergence Phase: Technology Empowering Core Business

类别涌现是指生产者及受众使用分类系统以外的组件和特征，在市场中形成、演化出的有价值的类别。数字平台生态系统的类别涌现，本质是技术优势驱动下突破既有分类系统、构建新价值类别的过程（Habtay，2012）。创新型企业作为有意愿颠覆既定秩序的市场力量，是推动类别涌现的直接动力。字节跳动通过数字技术优势构建，响应用户需求的协同逻辑，推动受众认知合法性塑造，在内容消费领域开创智能推荐驱动的新类别，实现从技术创新到生态位奠基的关键跨越。
Category emergence refers to the production and audience using components and features outside the classification system to form and evolve valuable categories in the market. The category emergence in the digital platform ecosystem is essentially a process of breaking through the existing classification system and building new value categories driven by technological advantages (Habtay, 2012). Innovative enterprises, as market forces willing to disrupt the existing order, are the direct driving force behind category emergence. ByteDance, through its technological advantages, constructs a collaborative logic responsive to user needs, promotes the shaping of audience cognitive legitimacy, and creates a new category driven by intelligent recommendations in the content consumption field, achieving a key leap from technological innovation to establishing an ecological niche.

4.1.1 数字技术涌现
Emergence of Digital Technology

类别涌现是指生产者及受众使用分类系统以外的组件和特征演化出有价值的类别（Durand & Khaire，2017）。一般而言，创新型企业作为有意愿颠覆既定秩序的市场力量，是推动类别涌现的直接动力（Montauti，2019）。在数字技术蓬勃发展的时代背景下，字节跳动旗下的今日头条、抖音等核心业务产品得以涌现。字节跳动聚焦于算法推荐系统与用户画像技术两项关键的数字技术组件，以协同过滤算法和深度学习模型为代表，这些技术被模块化封装，成为支撑产品运行的基石。协同过滤算法通过分析用户的行为数据，如浏览历史、点赞、评论等，发现具有相似兴趣爱好的用户群体，并为目标用户推荐该群体中其他人喜欢的内容。深度学习模型则能够自动学习数据中的复杂模式和特征，对用户的兴趣进行更精准的建模。二者相互配合，实现了内容的个性化分发，为用户提供“千人千面”的内容体验。随着业务的发展，字节跳动将算法推荐系统、用户画像技术与短视频、资讯等内容载体进行跨技术堆栈集成，实现了架构重组。以今日头条和抖音为代表，构建起以内容为入口的超级平台架构。在这个架构中，算法推荐技术成为连接用户与内容的核心纽带，它不仅负责将海量的内容进行筛选和分类，还根据用户画像为每个用户定制专属的内容流（Konstan & Riedl, 2012）。同时，用户在平台上的行为数据又不断反馈给算法，使其能够持续优化推荐效果，形成一个动态的、自我优化的循环。这种架构重组打破了传统媒体内容分发的模式，实现了内容的去中心化传播，无论是专业的内容创作者还是普通用户，都有机会通过优质内容获得大量曝光，激发了用户创作和参与的积极性，丰富了平台的内容生态。
Category emergence refers to the evolution of valuable categories by producers and audiences using components and features outside of classification systems (Durand & Khaire, 2017). Generally speaking, innovative enterprises, as market forces willing to disrupt the established order, are the direct driving force behind category emergence (Montauti, 2019). In the context of the vigorous development of digital technology, core business products such as Jinri Toutiao and Douyin under ByteDance have emerged. ByteDance focuses on two key digital technology components: algorithmic recommendation systems and user profile technology. Represented by collaborative filtering algorithms and deep learning models, these technologies are modularized and encapsulated, becoming the cornerstone of product operation. Collaborative filtering algorithms analyze user behavior data, such as browsing history, likes, and comments, to discover user groups with similar interests and recommend content liked by others in the group to target users. Deep learning models can automatically learn complex patterns and features from data, enabling more precise modeling of user interests. The two work together to achieve personalized content distribution, providing users with a "千人千面" content experience. As the business develops, ByteDance integrates algorithmic recommendation systems, user profile technology, and content carriers such as short videos and news across different technology stacks, achieving architectural restructuring. Represented by Jinri Toutiao and Douyin, they have built a super platform architecture with content as the entry point. In this architecture, algorithmic recommendation technology becomes the core link connecting users and content. It not only filters and categorizes massive amounts of content but also customizes a personalized content stream for each user based on user profiles (Konstan & Riedl, 2012). At the same time, user behavior data on the platform is continuously fed back to the algorithm, allowing it to continuously optimize recommendation effectiveness, forming a dynamic, self-optimizing cycle. This architectural restructuring breaks the traditional media content distribution model, achieving decentralized content dissemination. Whether professional content creators or ordinary users, everyone has the opportunity to gain extensive exposure through high-quality content, stimulating user creativity and participation, and enriching the platform's content ecosystem.

表2 数字平台生态系统类别创建阶段的典型证据
Table 2 Typical Evidence of the Creation Stage of the Digital Platform Ecosystem Category

4.1.2 跨类别创新战略
4.1.2 Cross-Category Innovation Strategy

（1）身份战略
(1) Identity Strategy

新类别在其形成的初期很少得到明确的定义，因此早期进入者对于类别的含义、边界甚至存在本身都存在很大的不确定性（Suarez et al.，2015）。要使一个类别得以存续并保持活力，至关重要的是要达到最佳的连贯性水平，需要内、外部视角下的身份均通过“认知基础设施”的检验。今日头条刚推出时，市场上已经有传统新闻客户端，用户可能对算法推荐感到新奇，但也有质疑，比如信息茧房的问题，早期接受者包括技术爱好者与信息需求型用户，如大学生、白领等。今日头条实施"信息熵平衡算法"，引入审核团队，通过嵌入微博及微信的分享功能使其与用户熟悉的场景相结合。初期保留“热点”板块，人工编辑内容占比30%，逐步过渡到纯算法推荐，使其更符合用户的认知过程。在抖音创立的过程中，创始人张一鸣在头条号创作者大会的演讲中，明确了All in短视频领域的口号，确立成为“年轻人的音乐短视频社区”的用户定位，以短视频为核心功能，主打15秒的炸裂音乐，符合年轻人的审美和节奏，定位了自身在分类系统中所占据的位置。区别于快手、美拍等依赖社交关系或热门标签的分发方式，抖音通过算法驱动的沉浸式体验快速精准地推荐给用户感兴趣的内容，通过算法推荐技术在用户心中构建了独特的新身份，让用户更容易发现感兴趣的内容，提高了用户粘性和使用时长。抖音通过春晚、综艺冠名等形式强化国民娱乐基础设施身份，提供了话题挑战、AR拍摄、模仿秀等春晚联动新玩法，形成了更为广泛的用户身份认知。
New categories rarely receive a clear definition in their early stages, so early entrants face significant uncertainty about the meaning, boundaries, and even the existence of the category itself (Suarez et al., 2015). For a category to endure and remain vibrant, it is crucial to achieve optimal coherence, requiring identities from both internal and external perspectives to pass the test of the "cognitive infrastructure." When Toutiao was first launched, the market already had traditional news clients, and users might find algorithmic recommendations novel, but there were also doubts, such as the issue of filter bubbles. Early adopters included tech enthusiasts and information-seeking users, such as college students and white-collar workers. Toutiao implemented the "information entropy balance algorithm," introduced an audit team, and integrated sharing functions from Weibo and WeChat to align with users' familiar scenarios. In the early stages, it retained a "hot topics" section with 30% manually edited content, gradually transitioning to pure algorithmic recommendations to better align with users' cognitive processes. During the creation of Douyin, founder Zhang Yiming, in his speech at the Toutiao Creator Conference, clearly stated the slogan of "All in short videos," establishing the user positioning as a "music short video community for young people." With short videos as the core function, it focused on explosive music lasting 15 seconds, aligning with the aesthetics and rhythm of young people, and defined its position in the classification system. Unlike platforms like Kuaishou and Meipai, which rely on social relationships or trending hashtags for distribution, Douyin quickly and accurately recommends content of interest to users through algorithm-driven immersive experiences. Through algorithmic recommendation technology, it constructs a unique new identity in users' minds, making it easier for them to discover content of interest and increasing user stickiness and usage time. Douyin reinforces its identity as a national entertainment infrastructure through events like the Spring Festival Gala and sponsored variety shows, offering new interactive features such as topic challenges, AR shooting, and imitation shows, thereby broadening user identity recognition.

（2）合法性战略
(2) Legitimacy Strategy

合法的类别成员资格并非只是抽象概念，而是具有实际价值的，分类规则为符合合法区域的类别提供了相对一致的理解和秩序框架，更符合受众的认知和期望，进而得到正向的评价和绩效，不在合法区域的类别会影响评价者的中介作用，导致他们减少关注或降低评级（Zuckerman，1999）。企业加入合法的类别会获得价值奖励，因此，组织需要通过各种策略来获取和维持合法性。根据2013年艾瑞咨询数据，42%的用户调查反馈担心今日头条的信息推荐系统会导致视野窄化，本研究中62.5%的访谈用户表示在最初接触时会担心信息茧房，同时会担心新闻的质量和真实性等问题。今日头条通过让不同年龄、地域、兴趣的用户都能找到有价值的内容，强调其信息的价值性来快速获取合法性；抖音迅速拓展涵盖生活、美食、知识科普、技能教学等多领域内容，打破传统娱乐边界，让大众认识到短视频不仅是娱乐消遣，还能成为知识获取、自我表达和社交互动的新途径。这种将新类别与用户认可的类别原型产生链接的合法性策略有助于减少受众的认知成本和不确定性（唐方成等，2023）。跨类别或被污名化的组织需要加强与合法形式的关联，可以通过转移污名，消减认知压力和负面评价的方式获得合法性（Delmestri & Greenwood，2016）。在类别涌现阶段，早期的算法对自媒体内容的质量控制不足，存在虚假信息传播、自媒体内容侵权、内容低俗、价值观偏差等问题频发，字节跳动作为创新企业不仅必须获取自身的合法性，还需要建立被社会接受的行业类别合法性（Reuber et al.，2017），积极与权威媒体、专业机构展开深度合作，借助其权威性和专业性，扶持优质内容创作者，搭建多层级、精细化的内容审核体系，重塑平台内容形象，引导行业正向发展，提升自身合法性。
Legal category membership is not just an abstract concept but has practical value. Classification rules provide a relatively consistent understanding and order framework for categories that fit legal regions, which aligns with the audience's cognition and expectations, thereby leading to positive evaluations and performance. Categories outside the legal region may affect the mediator's role of evaluators, causing them to reduce attention or lower ratings (Zuckerman, 1999). Enterprises joining legal categories receive value rewards, so organizations need to use various strategies to obtain and maintain legitimacy. According to iResearch data in 2013, 42% of user surveys were concerned about the information recommendation system of Jinri Toutiao leading to narrowed horizons. In this study, 62.5% of interview users expressed concerns about the filter bubble when they first encountered it, as well as worries about the quality and authenticity of news. Jinri Toutiao quickly gained legitimacy by allowing users of different ages, regions, and interests to find valuable content, emphasizing the value of its information. Douyin rapidly expanded to cover a wide range of content including daily life, food, science popularization, and skill teaching, breaking traditional entertainment boundaries and making the public recognize that short videos are not just for entertainment and leisure, but also new ways to acquire knowledge, self-expression, and social interaction. This legitimacy strategy that links new categories to user-recognized category prototypes helps reduce the audience's cognitive costs and uncertainty (Tang Fangcheng et al., 2023). Organizations that span categories or are stigmatized need to strengthen their association with legitimate forms, which can obtain legitimacy by transferring stigma, reducing cognitive pressure and negative evaluations (Delmestri & Greenwood, 2016). In the emergence stage of categories, early algorithms had insufficient quality control for self-media content, leading to frequent problems such as the spread of false information, infringement of self-media content, vulgar content, and value deviations. As an innovative enterprise, ByteDance not only needs to obtain the legitimacy of itself but also needs to establish the legitimacy of industry categories accepted by society (Reuber et al., 2017). It actively engages in in-depth cooperation with authoritative media and professional institutions, leveraging their authority and professionalism to support high-quality content creators, build a multi-level, refined content review system, reshape the platform's content image, guide the positive development of the industry, and enhance its legitimacy.

（3）独特性战略
(3) Uniqueness Strategy

类别涌现的新颖性源自于创新者构建和传达区别于现有类别的新物质属性，这一维度的竞争着重在于独特性和创新性。字节跳动产品依托强大的算法推荐技术，对用户的兴趣偏好、行为习惯进行深度分析，实现精准内容推荐。抖音算法会根据用户的浏览、点赞、评论等行为，精准推送符合其兴趣的视频，让用户不断刷到感兴趣的内容，沉浸其中。相比之下，部分同类短视频产品推荐的内容精准度不足，用户容易看到大量不感兴趣的信息，难以持续沉浸。今日头条也不仅局限于资讯阅读，还融合了问答、短视频等多种内容形式，满足用户不同场景下的信息获取和娱乐需求。抖音的用户定位最初主要是弥补短视频市场一二线城市的用户空白。客观来讲，抖音在进入短视频领域时，短视频的市场资源几乎被同类产品瓜分殆尽。当时快手的目标群体主要集中在我国三四线城市及大部分乡镇地区，从而使一二线城市成为快手目标群体的空白区域。个性和创意活动是符合一二线城市用户追求时尚潮流、喜欢挑战和创新的选择，抖音率先推出各种创新的特效、互动功能，如舞曲音乐视频，名画抖抖抖等创意视频。与此同时，抖音积极开拓海外业务推出国际化短视频平台Tiktok，收购Musical.ly并整合入TikTok，快速捕捉全球化短视频需求，体现对空白市场及市场需求的战略预判。字节跳动依托强大的算法推荐技术，树立自身独特的类别身份以在行业中立足，创造出更适应受众需求和市场规律的新类别。
The novelty of emerging categories stems from innovators creating and conveying new material properties that distinguish them from existing categories, with competition in this dimension focusing on uniqueness and innovation. ByteDance products rely on powerful algorithmic recommendation technology to deeply analyze users' interests and preferences, behavior habits, achieving precise content recommendations. Douyin's algorithm delivers videos that precisely match users' interests based on their browsing, likes, comments, etc., allowing users to continuously encounter content they are interested in and immerse themselves. In contrast, some similar short video products have insufficient content accuracy, making it easy for users to see a lot of uninteresting information and making it difficult to maintain immersion. Today's 头条 also goes beyond news reading, integrating various content forms such as Q&A and short videos to meet users' information acquisition and entertainment needs in different scenarios. Douyin's initial user positioning was mainly to fill the user gap in first and second-tier cities in the short video market. Objectively speaking, when Douyin entered the short video field, the market resources in the short video market had almost been exhausted by similar products. At that time, Kuaishou's target group was mainly concentrated in China's third and fourth-tier cities and most townships, making first and second-tier cities a blank area for Kuaishou's target group. Personality and creative activities align with the preferences of first and second-tier city users for fashion trends, challenges, and innovation. Douyin pioneered various innovative effects, interactive features, such as dance music videos, famous painting shake-shake videos, and other creative videos. Meanwhile, Douyin actively expanded overseas business by launching an international short video platform TikTok, acquiring Musical.ly and integrating it into TikTok, quickly capturing global short video demand, demonstrating strategic foresight for blank markets and market needs. ByteDance relies on its powerful algorithmic recommendation technology to establish a unique category identity in the industry and create a new category that better adapts to audience needs and market rules.

4.1.3 基于核心业务的数字平台生态系统
4.1.3 Digital Platform Ecosystem Based on Core Business

在数字平台生态系统涌现阶段的技术任务是奠定生态系统的基础设施，生态任务是塑造不同利益相关者在生态系统中扮演的角色，尽管这些角色部分由生态系统的技术架构定义，但需要通过定义行为规范和角色期望来强化这些角色（Autio，2021）。数字平台核心企业需要将算法推荐系统、用户画像技术进行模块化封装，形成可复用的技术组件，并通过跨技术堆栈集成，构建起以今日头条、抖音为核心的信息型平台生态系统。这种技术组件化策略不仅降低了内部业务拓展的技术成本，更重要的是为外部互补企业提供了标准化接入接口，使算法技术从业务支撑工具升级为生态基础设施。以抖音推荐算法为例，其从内容分发工具演变为决定内容传播路径、创作者机会和用户体验的核心机制，成为整个生态系统运行的技术中台，依托技术资源逐步构建起多边交互的生态系统，影响着生态系统的演化方向（Shipilov & Gawer，2020）。从平台与互补企业的角度来看，今日头条、抖音作为核心平台通过技术开放性和生态控制权平衡生态结构，提供协调机制、准入规则、知识产权和金融资本等制度及资源基础，吸引开发者、品牌商、MCN机构等互补企业加入，创造良好的用户体验，形成了价值共创网络。从用户角度来看，他们既是内容的消费者，也是内容的创造者。平台提供的便捷创作工具和广泛的传播渠道，吸引了大量用户参与到内容创作中，形成了丰富多样的内容生态。用户在消费内容的过程中，通过点赞、评论、分享等行为，为平台贡献了数据，这些数据又进一步优化了算法推荐，提升了内容分发的精准度。算法推荐从最初单纯的内容分发工具，逐渐升级为整个生态系统的基础设施，数字技术的作用从“业务支撑”成功转型为“生态定义”，字节跳动逐步构建起具有强大竞争力的信息型平台生态系统，建立基于核心业务的数字平台生态系统。
During the emergence phase of the digital platform ecosystem, the technological task is to lay the foundation for the infrastructure of the ecosystem. The ecological task is to shape the roles played by different stakeholders within the ecosystem. Although these roles are partly defined by the technological architecture of the ecosystem, they need to be reinforced through the definition of behavioral norms and role expectations (Autio, 2021). Core enterprises of digital platforms need to modularize and encapsulate algorithmic recommendation systems and user profiling technologies to form reusable technical components. Through cross-technology stack integration, they build an information-based platform ecosystem centered around platforms like Jinri Toutiao and Douyin. This technical componentization strategy not only reduces the technical costs of internal business expansion but more importantly provides standardized access interfaces for external complementary enterprises, upgrading algorithmic technology from a business support tool to an ecological infrastructure. Taking Douyin's recommendation algorithm as an example, it has evolved from a content distribution tool to a core mechanism determining the paths of content dissemination, creator opportunities, and user experience, becoming the technological backbone of the entire ecosystem. Relying on technological resources, it gradually builds a multi-stakeholder interactive ecosystem, influencing the evolutionary direction of the ecosystem (Shipilov & Gawer, 2020). From the perspective of the platform and complementary enterprises, platforms like Jinri Toutiao and Douyin balance the ecological structure through technological openness and ecological control rights. They provide institutional and resource foundations such as coordination mechanisms, entry rules, intellectual property, and financial capital to attract complementary enterprises like developers, brands, and MCN institutions to join, creating a good user experience and forming a value co-creation network. From the user's perspective, they are both consumers and creators of content. The convenient content creation tools and extensive dissemination channels provided by the platform have attracted a large number of users to participate in content creation, forming a rich and diverse content ecosystem. During the consumption of content, users contribute data through likes, comments, and shares, which in turn further optimizes the algorithmic recommendation, improving the accuracy of content distribution. Algorithmic recommendation has gradually evolved from a simple content distribution tool to the infrastructure of the entire ecosystem. The role of digital technology has successfully transitioned from "business support" to "ecosystem definition." ByteDance has gradually built a highly competitive information-based platform ecosystem, establishing a digital platform ecosystem based on core businesses.

4.2 类别跨越阶段：持续创新强化平台价值
4.2 Category Crossing Phase: Continuous Innovation Strengthening Platform Value

在类别跨越阶段，字节跳动通过技术—业务双轮驱动突破生态粘性与平台转换成本壁垒，构建起后台基础层、中台数据层与前台应用层的三元架构，将算法推荐、数据处理等通用技术能力沉淀至中台，实现跨业务场景的技术复用与快速迭代，显著降低创新成本（韩进等，2022）。通过技术架构与战略设计的耦合，形成技术深度嵌入和技术杠杆应用与生态协同相互强化，共同作用于平台的跨类别发展，实现生态边界的持续拓展与竞争优势的持续创新。
During the category crossing phase, ByteDance breaks through the barriers of ecological stickiness and platform conversion costs through dual driving of technology and business, constructing a ternary architecture of backend infrastructure layer, mid-layer data layer, and frontend application layer. It precipitates general technical capabilities such as algorithmic recommendation and data processing into the mid-layer to achieve technical reuse and rapid iteration across business scenarios, significantly reducing innovation costs (Han Jin et al., 2022). Through the coupling of technology architecture and strategic design, it forms a mutual reinforcement between deep technological embedding, technical leverage application, and ecological synergy, all of which act together on the platform's cross-category development, achieving continuous expansion of the ecological boundary and sustained innovation in competitive advantage.

4.2.1 数字技术跨类别应用
4.2.1 Cross-Category Application of Digital Technology

数字技术的跨类别应用是字节跳动突破战略刚性、实现持续创新的核心动能。面对伊卡洛斯悖论中效率导向流程可能引发的创新阻碍，字节跳动通过技术能力的模块化封装与跨业务赋能为平台业务提供快速、稳定的技术支持，开发团队无需为每个业务重新构建推荐系统，减少了技术研发的人力和时间投入，降低了创新的成本。字节跳动构建的“大中台+小前台”架构，中台层集成算法推荐、多模态内容理解、实时数据处理等核心技术，为前台业务提供标准化技术组件。例如，飞书作为企业协作平台，复用中台的用户画像与消息推送技术，同时保留独立的组织架构管理模块。复用成熟的数字技术能够加快新业务的开发和迭代速度，更推动了组织层面的认知升级（黄欣卓等，2022），通过分析抖音电商场景中用户的价格敏感度与内容消费场景的偏好差异，中台技术团队迭代出更精准的跨场景推荐模型，反哺内容分发与商品推荐效率。字节跳动的数据湖体系整合了内容消费、社交互动、电商交易等多源数据，通过用户行为标签、地理位置信息（POI）的跨场景关联，形成复合型需求图谱。以本地生活服务为例，抖音通过POI技术为线下商户添加地理位置标签，将用户浏览视频时的兴趣点与附近商家的团购信息匹配，实现内容浏览到一键购买的无缝衔接。这种共性技术统一供给、个性需求独立开发的模式，既确保了核心业务平台在算法底层逻辑的一致性，又允许跨类别业务平台根据垂直领域规则灵活适配，实现了战略统一性与业务独特性的动态平衡。
The cross-category application of digital technology is the core driving force for ByteDance to break through strategic rigidity and achieve continuous innovation. Facing the potential innovation barriers that efficiency-oriented processes in the Icarus paradox might cause, ByteDance provides rapid and stable technical support for platform businesses through modular encapsulation of technical capabilities and cross-business empowerment. Development teams do not need to rebuild recommendation systems for each business, reducing the human and time investment in R&D and lowering the cost of innovation. ByteDance's "big middle platform + small front platform" architecture integrates core technologies such as algorithmic recommendation, multimodal content understanding, and real-time data processing in the middle platform layer, providing standardized technical components for front-end businesses. For example, Feishu, as an enterprise collaboration platform, reuses the user profiling and message push technology of the middle platform while retaining an independent organizational structure management module. Reusing mature digital technology can accelerate the development and iteration speed of new businesses and further promote cognitive upgrades at the organizational level (Huang Xinzhuo et al., 2022). By analyzing users' price sensitivity and content consumption preferences in different scenarios in Douyin e-commerce, the middle platform's technical team iteratively developed a more accurate cross-scenario recommendation model, which in turn enhances the efficiency of content distribution and product recommendations. ByteDance's data lake system integrates multi-source data such as content consumption, social interaction, and e-commerce transactions. Through cross-scenario correlation of user behavior tags and geographical location information (POI), it forms a composite demand graph. Taking local life services as an example, Douyin uses POI technology to add geographical location tags to offline merchants, matching users' interest points when browsing videos with group purchase information from nearby merchants, achieving seamless connection from content browsing to one-click purchases. This model of providing common technology uniformly and developing personalized needs independently ensures consistency in the algorithmic logic at the core business platform while allowing cross-category business platforms to flexibly adapt to vertical domain rules, achieving a dynamic balance between strategic unity and business uniqueness.

数字技术跨类别应用并非简单的能力平移，需要结合多平台协同运作能力和业务场景持续优化（张翱、孙久文，2025）。字节跳动在将算法推荐技术从信息型平台迁移至交易型平台时，针对抖音电商商品交易的独特需求，如价格、库存、商家信誉等，增加了实时竞价、动态库存感知等模块，使基于用户行为数据模型推荐模型从内容相关性转向交易转化导向，帮助互补企业提升购买转化率。优化后的算法不仅提升了抖音电商的商品交易总额（GMV），其沉淀的实时决策机制又被应用于本地生活服务的商家排序，形成从通用技术能力到垂直场景适配，有效避免了单一领域技术专精化导致的创新僵化，使核心技术在跨类别创新中持续进化。火山引擎通过复用多模态内容理解技术，与上汽携手开发智能座舱交互系统，实现语音、手势等多维度的人机交互。技术外溢不仅强化了字节跳动的生态主导地位，更通过外部场景的反哺丰富了技术应用维度，拓展了在元宇宙、生成式人工智能（AIGC）等前沿领域，运用边缘计算与大模型技术的跨场景融合，推出了Pixsoul、豆包等跨类别创新平台，展现了字节跳动在创新型平台领域的探索潜力。
The cross-category application of digital technology is not a simple transfer of capabilities but requires the integration of multi-platform collaborative operation capabilities and continuous optimization of business scenarios (Zhang Ao, Sun Jiwen, 2025). When ByteDance migrates algorithmic recommendation technology from information-based platforms to transaction-based platforms, it adds modules such as real-time bidding and dynamic inventory perception to meet the unique needs of Douyin e-commerce product transactions, such as price, inventory, and merchant reputation. This shifts the user behavior data model-based recommendation model from content relevance to transaction conversion orientation, helping complementary businesses improve purchase conversion rates. The optimized algorithm not only increases the total transaction volume (GMV) of Douyin e-commerce but also applies its real-time decision-making mechanism to the merchant ranking of local life services, forming an effective adaptation from general technical capabilities to vertical scenarios, effectively avoiding the innovation rigidity caused by single-domain technology specialization, and allowing core technologies to continue evolving in cross-category innovation. Huoshan Engine reuses multimodal content understanding technology to collaborate with SAIC to develop an intelligent cockpit interaction system, achieving multi-dimensional human-computer interaction through voice, gestures, and other means. The spillover of technology not only strengthens ByteDance's ecological dominance but also enriches the dimensions of technology application through feedback from external scenarios, expanding its application in cutting-edge fields such as the metaverse and generative artificial intelligence (AIGC). By integrating edge computing and large model technologies across scenarios, ByteDance has launched cross-category innovation platforms such as Pixsoul and Daobao, demonstrating its exploration potential in innovative platform fields.

表3 数字平台生态系统类别跨越阶段的典型证据
Table 3 Typical Evidence of Cross-Category Phases in Digital Platform Ecosystems

4.2.2 跨类别创新战略
4.2.2 Cross-category innovation strategy

（1）平台包络战略
(1) Platform envelope strategy

平台包络战略将其自身平台的功能与目标平台功能融合或捆绑在一起，共享用户关系和通用组件，从而形成更为多功能的平台，以此扩展自身平台的市场边界（Eisenmann et al.，2011）。通常指平台通过扩展相邻业务，利用现有资源和用户基础，覆盖更多市场。字节跳动在数字平台生态系统构建过程中，充分运用平台包络战略实现跨类别创新，极大地拓展了市场边界。抖音基于庞大的用户基础和强大的数据流量，从短视频领域向内容消费领域跨界，在产业链上进行纵向包络，抖音主页面嵌入商城、团购、直播等跨类别业务功能入口，用户浏览美妆视频时可一键跳转商品链接完成购买，观看探店视频时能实时领取附近商家的团购券，符合用户对数字平台类别组合能够连贯认知的新形态，满足了用户在单一平台使用多平台协同服务的新需求。2023年抖音电商GMV突破2万亿元，本地生活服务覆盖全国450万家门店。字节跳动将番茄小说、红果短剧等内容平台纳入统一技术与数据中台，将智力产权（IP）资源进行跨类别开发，实现横向的内容生态包络。例如，番茄小说的热门网文通过AI生成短视频剧情片段，在抖音进行引流推广；红果短剧的爆款内容反向带动原著小说阅读量增长。这种横向包络不仅丰富了抖音的内容供给，更通过各平台间的用户体系互通，打破单一短视频平台的内容局限，形成全形态内容创作生态。
The platform envelopment strategy integrates or bundles the functions of its own platform with those of the target platform, sharing user relationships and common components, thereby forming a more multifunctional platform to expand the market boundaries of its own platform (Eisenmann et al., 2011). It usually refers to the platform expanding adjacent businesses by utilizing existing resources and user bases to cover more markets. ByteDance fully leverages the platform envelopment strategy in the process of building its digital platform ecosystem to achieve cross-category innovation, greatly expanding market boundaries. Douyin, based on its vast user base and strong data traffic, has expanded from the short video field to the content consumption field, conducting vertical envelopment in the industrial chain. Douyin's main page embeds cross-category business function entrances such as shopping malls, group purchases, and live streaming. Users can jump to product links with one click while browsing beauty videos, and can instantly redeem group purchase coupons from nearby merchants while watching store exploration videos, which aligns with the new form of users' continuous perception of category combinations in digital platforms and meets the new demand for using multi-platform collaborative services on a single platform. In 2023, Douyin's e-commerce GMV exceeded 2 trillion yuan, and local life services covered 4.5 million stores nationwide. ByteDance has included content platforms such as 番茄小说 and 红果短剧 into a unified technology and data platform, developing intellectual property (IP) resources across categories, and achieving horizontal content ecosystem envelopment. For example, popular web novels from 番茄小说 are used to generate short video plot segments using AI for promotion on Douyin; the hit content of 红果短剧 has driven an increase in the reading volume of the original novels. This horizontal envelopment not only enriches Douyin's content supply but also breaks the content limitations of a single short video platform through inter-platform user system interconnection, forming a full-form content creation ecosystem.

（2）开放式创新战略
(2) Open Innovation Strategy

字节跳动的跨类别创新战略不仅依赖内部技术突破，更积极践行开放式创新战略。通过多维度、多层次的创新协作与资源整合，塑造了独特的竞争优势，驱动业务的多元化拓展与持续增长（刘海兵，2019）。开放式创新战略通过技术接口开放与外部资源整合，平台核心企业通过API开放度控制生态边界，构建多边参与的创新生态（Gawer & Cusumano，2014）。抖音开放了视频创作、内容推荐、用户互动等多个层面的API，开发者基于这些接口开发出各类创新应用，涵盖特效滤镜、互动游戏、内容管理工具等领域，这不仅丰富了抖音的功能生态，满足用户多样化需求，还为字节跳动带来了源源不断的创意与技术。开放小程序接口接入美团、携程等服务，用户可在抖音内直接完成酒店预订、外卖点餐，进一步丰富了平台的应用场景。字节跳动同时运用自身的技术优势对外开放赋能，通过构建创作者生态协同网络，将云计算与AI能力向外部企业开放，将内部技术能力转化为行业共性解决方案，运用数据分析和智能推荐技术服务拓宽了技术应用边界，吸引外部企业加入，共同提升技术水平和业务能力，同时也为字节跳动带来了更多的合作机会和商业价值，形成了内外部协同创新网络，构建多边参与、价值共享的创新生态系统，实现跨行业资源整合与场景拓展。
ByteDance's cross-category innovation strategy not only relies on internal technological breakthroughs but also actively practices the open innovation strategy. Through multi-dimensional and multi-level innovation collaboration and resource integration, it has shaped unique competitive advantages, driving diversified business expansion and continuous growth (Liu Haibing, 2019). The open innovation strategy integrates external resources through open technology interfaces, and the core platform enterprises control the ecological boundaries through the openness of APIs, constructing an innovation ecosystem with multi-party participation (Gawer & Cusumano, 2014). Douyin has opened APIs at multiple levels such as video creation, content recommendation, and user interaction, allowing developers to create various innovative applications based on these interfaces, covering areas such as special effects filters, interactive games, and content management tools. This not only enriches Douyin's functional ecosystem and meets diverse user needs but also brings ByteDance a continuous stream of creativity and technology. Douyin has also opened mini-program interfaces to integrate services such as Meituan and Ctrip, allowing users to complete hotel bookings and take-out orders directly within Douyin, further enriching the platform's application scenarios. ByteDance also uses its technological advantages to empower external parties by building a collaborative network for creator ecosystems, opening cloud computing and AI capabilities to external enterprises, transforming internal technological capabilities into industry common solutions. It has used data analysis and intelligent recommendation services to broaden the boundaries of technology application, attracting external enterprises to join and jointly improve technological and business capabilities, while also bringing more cooperation opportunities and commercial value to ByteDance. This has formed an internal and external collaborative innovation network, constructing an innovation ecosystem with multi-party participation and value sharing, achieving cross-industry resource integration and scenario expansion.

（3）地位战略
(3) Positioning Strategy

随着组织从混沌走向有序，类别从原型走向跨越使得类别包含的范围不断扩大，类别对现有元素和属性重新排列、重新解释和重新评估，通过不断引入战略要素拓展了新的资源和机会，核心类别企业的地位不断攀升，处于优势生态位的主导地位类别其获取社会、文化和物质资源的能力将会更强。字节跳动早期凭借信息资讯类产品今日头条，在移动互联网内容分发领域崭露头角。此后，公司敏锐捕捉到短视频领域的巨大潜力，成功打造出抖音和TikTok等现象级产品。随着抖音用户规模的爆发式增长，字节跳动进一步挖掘用户需求和市场潜力，将业务拓展至电商领域。在跨越不同业务类别的发展过程中，抖音以短视频和直播为载体，结合庞大的用户流量，构建起电商业务新生态。抖音充分利用在短视频领域积累的主导地位类别优势，为电商业务提供清晰的市场定义和边界。一方面，抖音借助短视频和直播的形式，为商家提供了全新的商品展示和推广渠道，吸引大量商家入驻，丰富了商品品类；另一方面，通过大数据分析用户偏好，实现精准的商品推荐，提升用户购物体验。企业基于主导地位类别来定位自己的产品和技术方向更有助于获取机会窗口，可以利用主导地位类别提供的市场定义和边界，进行有针对性的技术实验和市场开发，通过持续的创新和适应市场需求，逐渐在其领域内建立起技术优势和市场地位（Suarez et al.，2015）。
As organizations transition from chaos to order, categories evolve from prototypes to cross-cutting, expanding the scope of categories, and rearranging, reinterpreting, and re-evaluating existing elements and attributes. By continuously introducing strategic elements, new resources and opportunities are expanded, and the status of core category enterprises continues to rise. Category enterprises in dominant ecological niches will have stronger abilities to acquire social, cultural, and material resources. In the early days, ByteDance relied on its information and news product Jinri Toutiao to gain prominence in the mobile internet content distribution field. Thereafter, the company keenly detected the huge potential in the short video field and successfully created phenomenal products such as Douyin and TikTok. As Douyin's user base experienced explosive growth, ByteDance further explored user needs and market potential, expanding its business into the e-commerce field. In the process of crossing different business categories, Douyin, using short videos and live streaming as carriers, combined with its vast user traffic, has built a new ecological system for e-commerce. Douyin fully utilizes its dominant position category advantage in the short video field to provide a clear market definition and boundaries for the e-commerce business. On one hand, Douyin, through the form of short videos and live streaming, provides merchants with brand new product display and promotion channels, attracting a large number of merchants to join and enriching the product categories; on the other hand, by analyzing user preferences through big data, it achieves precise product recommendations, enhancing the user shopping experience. Enterprises, based on the dominant position category, to position their own products and technology direction, is more helpful to obtain opportunity windows. They can utilize the market definition and boundaries provided by the dominant position category for targeted technological experiments and market development, through continuous innovation and adaptation to market demand, gradually establishing technical advantages and market status in their field (Suarez et al., 2015).

4.2.3 基于跨类别协同的数字平台生态系统
4.2.3 Digital Platform Ecosystem Based on Cross-Category Collaboration

跨类别协同的数字平台生态系统是技术跨类别应用驱动下的多维价值构建。字节跳动通过技术架构创新与战略设计耦合，将短视频、电商、本地生活、云服务与AIGC等跨类别业务整合成有机生态，展现了从单一信息型生态向跨类别价值共创生态的演进路径（Cusumano et al., 2019）。数字技术跨类别应用让数字平台企业能够跨行业竞争，构建起后台基础层、中台数据层与前台应用层的三元架构，各层业务紧密关联、协同运作，通过平台包络战略将新的功能纳入到核心平台中，实现跨类别场景融合与创新，促进效率性协同和价值性协同的产生（葛安茹、唐方成，2021）。数字平台企业结合现有资源和用户基础，权衡平台包络战略的范围，承担认知边界拓展、平衡制度逻辑冲突、跨类别合法性获取等职能，使创新类别产品渗透市场，拓宽社会文化和企业战略规划的空间，创造结合用户广泛意图、更具创新的跨类别数字平台生态系统。商家、MCN机构等互补企业纷纷嵌入数据流量闭环，通过数据共享、资源共享和技术协同，推动整个生态系统的演进。其中，平台向外部企业开放资源及能力，内容生产者提供高质量的内容，开发者提供互补技术与创意，支付平台处理交易，硬件制造商提供设备支持，多方共同实现了资源共享和优势互补，满足用户的需求并释放多元化的商业价值，再通过用户数据与互补企业资源的交叉复用，降低运营成本，提升资源配置效率，促进跨类别业务的协同发展。基于跨类别协同的数字平台生态系统能够精准定位技术方向，获取资源与技术优势，践行开放式创新战略，通过技术接口开放、外部资源整合，构建多边创新生态，最终形成用户规模扩大—互补企业收益增加—平台跨类别创新的正反馈循环。
A cross-category collaborative digital platform ecosystem is a multi-dimensional value construction driven by cross-category application of technology. ByteDance integrates cross-category businesses such as short videos, e-commerce, local life services, cloud services, and AIGC through technological architecture innovation coupled with strategic design, forming an organic ecosystem, demonstrating the evolutionary path from a single information-based ecosystem to a cross-category value co-creation ecosystem (Cusumano et al., 2019). Cross-category application of digital technology enables digital platform enterprises to compete across industries, constructing a ternary architecture of backend infrastructure layer, mid-layer data layer, and frontend application layer. Each layer's business is closely linked and collaboratively operated. Through platform envelopment strategy, new functions are integrated into the core platform, achieving cross-category scenario integration and innovation, promoting the generation of efficiency-based collaboration and value-based collaboration (Ge Anru, Tang Fangcheng, 2021). Digital platform enterprises, leveraging existing resources and user base, weigh the scope of platform envelopment strategy, undertake functions such as expanding cognitive boundaries, balancing institutional logic conflicts, and acquiring cross-category legitimacy, enabling innovative category products to penetrate the market, broadening the space for social culture and corporate strategic planning, creating a cross-category digital platform ecosystem that combines broad user intentions and greater innovation. Merchants, MCN institutions, and complementary enterprises 纷纷 embed themselves into the data flow loop, driving the evolution of the entire ecosystem through data sharing, resource sharing, and technical collaboration. Among them, the platform opens up resources and capabilities to external enterprises, content producers provide high-quality content, developers offer complementary technologies and creativity, payment platforms handle transactions, and hardware manufacturers provide device support. All parties achieve resource sharing and complementary advantages, meeting user needs and releasing diversified commercial value. Through the cross-use of user data and complementary enterprise resources, operational costs are reduced, resource allocation efficiency is improved, and collaborative development of cross-category businesses is promoted. Based on cross-category collaborative digital platform ecosystems, technology direction can be accurately identified, resource and technological advantages can be obtained, and open innovation strategies can be practiced. Through open technology interfaces and external resource integration, a multi-party innovation ecosystem is constructed, ultimately forming a positive feedback loop of user scale expansion—complementary enterprise revenue increase—platform cross-category innovation.

综上所述，数字平台生态系统的战略选择与发展演化是数字化时代管ß理研究的重要议题。本研究基于分类理论，按照“技术驱动—战略行动—生态演化”的研究框架探究了跨类别创新战略与数字平台生态系统的演化过程，具体模型如图4所示。
In summary, the strategic choices and evolutionary process of digital platform ecosystems are important research topics in digital age management studies. This study, based on classification theory and following the research framework of "technology-driven—strategic action—ecological evolution," explores the evolutionary process of cross-category innovation strategy and digital platform ecosystems, with the specific model shown in Figure 4.

图4 数字平台生态系统跨类别创新战略及演化模型
Figure 4 Cross-category innovation strategy and evolutionary model of digital platform ecosystem

5. 研究结论与展望
5. Research Conclusions and Outlook

5.1 研究结论
5.1 Research Conclusions

本研究以分类理论为基础，通过纵向案例研究揭示了数字平台生态系统跨类别创新的阶段性战略选择与演化过程，主要结论如下：
This study, based on classification theory, reveals the phased strategic choices and evolutionary process of cross-category innovation in digital platform ecosystems through longitudinal case studies. The main conclusions are as follows:

（1）数字技术对跨类别创新战略的驱动作用。现代社会正处于一个不断变革的时期，在类别涌现时期，核心技术是数字平台生态系统发展的基石，技术驱动表现为核心技术的模块化封装与用户认知重构，通过身份战略、合法性战略与独特性战略构建核心生态系统，其中，类别认知在跨类别创新战略中发挥关键引导作用，平台对用户认知规则的敏锐洞察是识别创新方向的前提，通过跨类别身份战略与合法性战略，能够突破用户固有的认知边界，在激烈的市场竞争中确立独特的平台定位。随着用户需求多元化，跨类别创新创造的独特体验，能够更好地满足人们多样化的期望。数字平台企业依据上述规则制定战略决策，提前布局新业务，提供更具的专业能力的产品，使新类别满足市场竞争合法性和用户体验独特性。在类别跨越阶段，平台包络战略为用户提供了更为便捷的一站式服务，通过将不同类型的业务包络至核心平台，有助于核心平台保持领先地位。技术与业务的双轮驱动促进跨类别构建开放式的平台架构，通过快速迭代产品和技术，及时调整研发方向与业务策略是获得战略优势的方向。技术驱动的跨类别创新需要持续投入大量资源进行基础研发、业务模式创新和用户体验优化，推出新的功能和产品，以应对快速变化的竞争环境，不断巩固和拓展市场份额。
(1) The driving role of digital technology in cross-category innovation strategies. Modern society is in a period of constant transformation. During the emergence of categories, core technology is the cornerstone of the development of digital platform ecosystems. The technological driver is manifested in the modular encapsulation of core technologies and the reconstruction of user cognition. Through identity strategy, legitimacy strategy, and uniqueness strategy, core ecosystems are constructed. Among them, category cognition plays a key guiding role in cross-category innovation strategies. The platform's keen insight into user cognitive rules is the prerequisite for identifying innovation directions. Through cross-category identity strategy and legitimacy strategy, it can break through users' inherent cognitive boundaries and establish a unique platform positioning in fierce market competition. As user needs diversify, the unique experiences created by cross-category innovation can better meet people's diverse expectations. Digital platform enterprises formulate strategic decisions based on the above rules, lay out new businesses in advance, and provide products with greater professional capabilities, so that new categories meet market competition legality and user experience uniqueness. In the category crossing stage, the platform envelope strategy provides users with more convenient one-stop services. By enveloping different types of businesses into the core platform, it helps the core platform maintain its leading position. The dual drive of technology and business promotes the construction of an open platform architecture in cross-category. By rapidly iterating products and technology and timely adjusting R&D direction and business strategies, it is the direction to gain strategic advantages. Technologically driven cross-category innovation requires continuous investment of a large amount of resources in basic R&D, business model innovation, and user experience optimization, launching new features and products to respond to the rapidly changing competitive environment, and continuously consolidating and expanding market share.

（2）数字平台生态系统跨类别创新战略的独特优势。在跨类别创新过程中，数字平台企业将在特定领域积累的技术能力、数据洞察与用户基础等核心优势迁移至新类别，塑造独特的价值主张，避免同质化竞争，使平台在跨类别业务中快速建立技术壁垒，实现业务和市场的拓展。当数字平台拥有海量的用户基础时，将能吸引更多的互补企业及用户入驻，形成规模优势和网络效应，为数字平台带来了更多的商业机会，使其能够在跨类别业务中实现快速发展，进入到竞争对手的生存空间。与此同时，企业需在保持核心技术逻辑与用户价值主张一致性的同时，针对不同类别业务特性进行适应性创新，确保组织内部结构与外部生态关系的连贯性。在数字平台生态系统中，协同合作是实现跨类别创新的重要基础。数字平台企业与生态系统内的互补者建立紧密的合作关系，通过整合各方资源，实现优势互补，共同推动不同业务类别之间的融合与创新。成功的生态系统应使参与者对其位置感到满意，数字平台企业需要在协同合作与竞争之间找到战略平衡，根据自身的市场环境，灵活调整战略。在类别创建阶段加强与互补企业的合作，共同开拓市场；在类别跨越领域，通过持续创新的目标和技术深度、技术杠杆的驱动提升自身创新能力和主导地位。
(2) The unique advantages of cross-category innovation strategies in digital platform ecosystems. In the process of cross-category innovation, digital platform enterprises will migrate core advantages such as technical capabilities, data insights, and user bases accumulated in specific fields to new categories, shaping unique value propositions, avoiding homogenization competition, and enabling platforms to quickly establish technical barriers in cross-category businesses, achieving business and market expansion. When digital platforms have a massive user base, they can attract more complementary enterprises and users to join, forming scale advantages and network effects, which brings more business opportunities to digital platforms, enabling them to achieve rapid development in cross-category businesses and enter the survival space of competitors. At the same time, enterprises need to maintain consistency in the logic of core technology and user value propositions while innovating adaptively for the characteristics of different category businesses, ensuring the continuity of internal organizational structure and external ecological relationships. In the digital platform ecosystem, collaboration is the important foundation for cross-category innovation. Digital platform enterprises establish close cooperative relationships with complementary parties within the ecosystem, integrate resources from all parties, achieve complementary advantages, and jointly promote the integration and innovation between different business categories. A successful ecosystem should make participants feel satisfied with their position. Digital platform enterprises need to find a strategic balance between collaboration and competition, flexibly adjusting strategies according to their market environment. Strengthen cooperation with complementary enterprises in the category creation stage to jointly explore the market; in the category crossing field, continuously improve innovation capabilities and dominant position through the drive of continuous innovation goals, technological depth, and technological leverage.

（3）数字平台生态系统跨类别创新战略与演化的共演机制。数字平台生态系统的跨类别创新与演化呈现双向互动的共演机制，核心体现为技术驱动、战略行动与生态系统的动态耦合，共同成为推动数字平台生态系统演化发展的重要力量。跨类别创新战略以技术驱动为底层逻辑，通过核心技术的模块化沉淀与跨类别复用，为生态演化提供持续动能。在此基础上，通过身份构建、平台包络、开放式创新等战略行动，突破传统行业分类边界，驱动用户认知迁移与市场分类重构，战略行动促使数字资源与新类别业务深度整合，推动生态系统从初始的信息型生态系统，逐步演化为涵盖交易型、创新型的跨类别生态系统，实现技术能力的跨场景溢出与价值创造。在组织结构层面，平台从简单业务架构演变为多层级、多维度的复杂系统，不同业务板块通过技术接口与数据共享实现有机联动；在价值体系层面，单一主体的线性价值创造模式转向多元主体参与的网状价值共创模式，平台、互补企业、开发者、用户等通过合作互动实现资源整合与价值增值。这种演化不仅拓展了平台的战略空间，更通过场景融合与生态协同为新的跨类别创新提供了丰富的应用场景与资源支撑。数字平台生态系统的共演机制本质是技术、战略与生态的系统性协同，跨类别创新战略既是生态演化的驱动力，也在生态升级中获取新的发展机遇，最终形成可持续的创新生态。企业需把握技术驱动下的生态结构转型机遇，构建多元价值共创体系，以持续跨类别探索激发创新活力，以存续的竞争能力保障战略韧性，实现从基于核心业务到跨类别协同生态的跨越式发展。
(3) Co-evolution mechanism of cross-category innovation strategies and evolution in the digital platform ecosystem. The cross-category innovation and evolution of the digital platform ecosystem exhibit a bidirectional interactive co-evolution mechanism, primarily characterized by technology-driven, strategic actions, and dynamic coupling of the ecosystem, collectively becoming a significant force in promoting the evolution and development of the digital platform ecosystem. The cross-category innovation strategy, with technology driving as its underlying logic, provides continuous momentum for ecosystem evolution through modularization and cross-category reuse of core technologies. On this basis, strategic actions such as identity construction, platform envelopment, and open innovation break the boundaries of traditional industry classifications, drive the migration of user cognition and the restructuring of market categories, and facilitate the deep integration of digital resources and new category businesses. This pushes the ecosystem to evolve from an initial information-based ecosystem to a cross-category ecosystem encompassing transactional and innovative types, achieving cross-scenario spillover of technological capabilities and value creation. At the organizational structure level, the platform evolves from a simple business architecture into a multi-tiered, multi-dimensional complex system, with different business segments organically linked through technical interfaces and data sharing. At the value system level, the linear value creation model of a single entity shifts to a web-based value co-creation model involving multiple stakeholders, including platforms, complementary enterprises, developers, and users, who achieve resource integration and value enhancement through cooperative interactions. This evolution not only expands the strategic space of the platform but also provides rich application scenarios and resource support for new cross-category innovation through scenario integration and ecosystem synergy. The co-evolution mechanism of the digital platform ecosystem is essentially the systematic synergy of technology, strategy, and the ecosystem. The cross-category innovation strategy is not only a driver of ecosystem evolution but also gains new development opportunities during ecosystem upgrading, ultimately forming a sustainable innovation ecosystem. Enterprises need to seize the opportunities of structural transformation driven by technology, build a multi-stakeholder value co-creation system, and continuously explore cross-category innovation to stimulate innovation vitality, ensuring strategic resilience through sustained competitive capabilities, and achieving a leapfrog development from core business-based to cross-category collaborative ecosystems.

5.2 管理意义
5.2 Management Significance

针对现有研究在数字平台企业跨类别创新动态演化过程剖析不足的问题，本研究以字节跳动为案例，深入探究不同阶段跨类别创新战略行动以及生态系统演化发展。在理论层面，首次系统地总结出跨类别创新理论模型，选取了分类理论作为平台创新战略、平台发展变革分析的理论基础，整合身份、合法性、独特性等战略，将技术创新、市场认知和组织变革纳入统一分析框架，填补了关于组织行动与能力的分类战略仍然较少的研究缺口，为透视数字平台生态系统结构本质及战略选择适配提供了新的研究路径。在实践层面，研究为数字平台企业制定跨类别创新战略提供了极具价值的参考，企业可依据本研究提出的模型，结合自身发展阶段和市场环境，识别创新情境，培育关键驱动能力，合理规划战略行动，平衡跨界程度，把握生态系统变化趋势，有助于提升跨类别创新的成功率和竞争力，为企业构建持续创新的生态系统提供新的思路，促进数字平台生态系统在复杂多变的环境中实现突破和创新。
Addressing the insufficient analysis of the dynamic evolution process of cross-category innovation in digital platform enterprises in existing research, this study takes ByteDance as a case to deeply explore cross-category innovation strategies at different stages and the evolution of the ecosystem. At the theoretical level, it systematically summarizes the theoretical model of cross-category innovation for the first time, selecting classification theory as the theoretical foundation for platform innovation strategy and platform development transformation analysis. It integrates strategies such as identity, legitimacy, and uniqueness, incorporating technological innovation, market perception, and organizational change into a unified analytical framework, filling the gap in research on the classification strategies of organizational actions and capabilities, and providing a new research path for understanding the essence of the structure of digital platform ecosystems and the adaptation of strategic choices. At the practical level, the study offers valuable references for digital platform enterprises to formulate cross-category innovation strategies. Enterprises can use the model proposed in this study, combined with their own development stage and market environment, to identify innovation contexts, cultivate key driving capabilities, rationally plan strategic actions, balance cross-boundary extent, grasp ecosystem change trends, and enhance the success rate and competitiveness of cross-category innovation. This helps in building a continuously innovative ecosystem for enterprises and promotes the breakthrough and innovation of digital platform ecosystems in complex and changing environments.

5.3 研究局限与展望
5.3 Research Limitations and Prospects

本研究在探究数字平台生态系统跨类别创新战略及演化方面取得一定成果，但仍存在局限性。一方面，对持续创新能力在数字平台场景的应用讨论不够充分。虽然识别出多种跨类别创新战略，但持续创新能力在不同场景下的具体应用方式、影响因素及作用机制尚未深入剖析，后续研究可通过更多实证研究进行细化。另一方面，战略是复杂的概念，本研究虽已提出相关理论模型，但对战略制定和实施过程中的细节，如战略决策的动态调整、不同战略间的协同等研究不足。与此同时，还需要关注数字平台生态系统整体的健康发展，制定合理的制度与规则，确保各方利益得到平衡，实现生态系统的长期稳定发展。未来可深入研究创新战略与生态协同演化的影响因素和实现路径，开展更为细致的讨论。
This study has achieved certain results in exploring the cross-category innovation strategies and evolution of digital platform ecosystems, but there are still limitations. On one hand, the discussion on the application of continuous innovation capabilities in the context of digital platforms is not sufficient. Although various cross-category innovation strategies have been identified, the specific application methods, influencing factors, and mechanisms of continuous innovation capabilities in different scenarios have not been thoroughly analyzed. Future research can refine these aspects through more empirical studies. On the other hand, strategy is a complex concept, and although this study has proposed a relevant theoretical model, there is insufficient research on the details of strategy formulation and implementation, such as the dynamic adjustment of strategic decisions and the synergy between different strategies. At the same time, it is also necessary to focus on the overall healthy development of digital platform ecosystems, formulating reasonable systems and rules to ensure the balance of interests among all parties and achieve the long-term stable development of the ecosystem. Future research can delve deeper into the influencing factors and implementation paths of innovation strategies and ecosystem co-evolution, conducting more detailed discussions.

参考文献
References

Hein A, Schreieck M, Riasanow T, et al. Digital platform ecosystems[J]. Electronic markets, 2020, 30: 87-98.

Miller D. Organizational configurations: Cohesion, change, and prediction[J]. Human relations, 1990, 43(8): 771-789.

Navis C, Glynn M A. How new market categories emerge: Temporal dynamics of legitimacy, identity, and entrepreneurship in satellite radio, 1990–2005[J]. Administrative science quarterly, 2010, 55(3): 439-471.

唐方成, 靳晓曼, 顾世玲. 评价者分类视角下商业模式创新的合法化机制——以网约车的类别涌现为例[J]. 管理世界, 2023, 39(8): 114-131.
Tang Fangcheng, Jin Xiaoman, Gu Shiling. Legitimization mechanisms of business model innovation from the perspective of evaluator categorization - taking the emergence of categories of ride-hailing as an example[J]. Management World, 2023, 39(8): 114-131.

Kennedy M T, Fiss P C. An ontological turn in categories research: From standards of legitimacy to evidence of actuality[J]. Journal of Management studies, 2013, 50(6): 1138-1154.

Zuckerman E W. The categorical imperative: Securities analysts and the illegitimacy discount[J]. American journal of sociology, 1999, 104(5): 1398-1438.

Hsu G, Hannan M T. Identities, genres, and organizational forms[J]. Organization science, 2005, 16(5): 474-490.

Phillips D J, Zuckerman E W. Middle-status conformity: Theoretical restatement and empirical demonstration in two markets[J]. American Journal of Sociology, 2001, 107(2): 379-429.

Durand R, Paolella L. Category stretching: Reorienting research on categories in strategy, entrepreneurship, and organization theory[J]. Journal of Management Studies, 2013, 50(6): 1100-1123.

Zuckerman E W. The categorical imperative revisited: Implications of categorization as a theoretical tool[M]//From categories to categorization: Studies in sociology, organizations and strategy at the crossroads. Emerald Publishing Limited, 2017: 31-68.

Pontikes E G. Two sides of the same coin: How ambiguous classification affects multiple audiences’ evaluations[J]. Administrative Science Quarterly, 2012, 57(1): 81-118.

Allen R T, McDonald R M. Methodological pluralism and innovation in data-driven organizations[J]. Administrative Science Quarterly, 2025: 00018392251313737.

Durand R, Khaire M. Where do market categories come from and how? Distinguishing category creation from category emergence[J]. Journal of Management, 2017, 43(1): 87-110.

Gawer A, Cusumano M A. Industry platforms and ecosystem innovation[J]. Journal of product innovation management, 2014, 31(3): 417-433.

Kwak K, Kim W, Park K. Complementary multiplatforms in the growing innovation ecosystem: Evidence from 3D printing technology[J]. Technological Forecasting and Social Change, 2018, 136: 192-207.

Eisenmann T, Parker G, Van Alstyne M. Platform envelopment[J]. Strategic management journal, 2011, 32(12): 1270-1285.

Cudennec A, Durand R. Valuing spanners: Why category nesting and expertise matter[J]. Academy of Management Journal, 2023, 66(1): 335-365.

Glynn M A, Navis C. Categories, identities, and cultural classification: Moving beyond a model of categorical constraint[J]. Journal of Management Studies, 2013, 50(6): 1124-1137.

Ocasio W, Loewenstein J, Nigam A. How streams of communication reproduce and change institutional logics: The role of categories[J]. Academy of Management Review, 2015, 40(1): 28-48.

Walrave B, Talmar M, Podoynitsyna K S, et al. A multi-level perspective on innovation ecosystems for path-breaking innovation[J]. Technological forecasting and social change, 2018, 136: 103-113.

Benitez G B, Ghezzi A, Frank A G. When technologies become Industry 4.0 platforms: Defining the role of digital technologies through a boundary-spanning perspective[J]. International Journal of Production Economics, 2023, 260: 108858.

Pershina R, Soppe B, Thune T M. Bridging analog and digital expertise: Cross-domain collaboration and boundary-spanning tools in the creation of digital innovation[J]. Research Policy, 2019, 48(9): 103819.

Cennamo C. Competing in digital markets: A platform-based perspective[J]. Academy of Management Perspectives, 2021, 35(2): 265-291.

Mann G, Karanasios S, Breidbach C F. Orchestrating the digital transformation of a business ecosystem[J]. The Journal of Strategic Information Systems, 2022, 31(3): 101733.

Yin Y, Stecke K E, Li D. The evolution of production systems from Industry 2.0 through Industry 4.0[J]. International Journal of Production Research, 2018, 56(1-2): 848-861.

Eisenhardt K M. Building theories from case study research[J]. Academy of management review, 1989, 14(4): 532-550.

苏敬勤,王娜,高昕,等.案例学的构建——学理基础与现实可行性[J].管理世界,2021,37(09):207-214.DOI:10.19744/j.cnki.11-1235/f.2021.0147.
Su Jingqin, Wang Na, Gao Xin, et al. The construction of case study theory: Theoretical foundation and practical feasibility[J]. Management World, 2021, 37(09): 207-214. DOI: 10.19744/j.cnki.11-1235/f.2021.0147.

Gioia D A, Corley K G, Hamilton A L. Seeking qualitative rigor in inductive research: Notes on the Gioia methodology[J]. Organizational research methods, 2013, 16(1): 15-31.

Habtay S R. A firm‐level analysis on the relative difference between technology‐driven and market‐driven disruptive business model innovations[J]. Creativity and Innovation Management, 2012, 21(3): 290-303.

Durand R, Khaire M. Where do market categories come from and how? Distinguishing category creation from category emergence[J]. Journal of Management, 2017, 43(1): 87-110.

Montauti M, Wezel F C. Charting the territory: Recombination as a source of uncertainty for potential entrants[J]. Organization Science, 2016, 27(4): 954-971.

Konstan J A, Riedl J. Recommender systems: from algorithms to user experience[J]. User modeling and user-adapted interaction, 2012, 22: 101-123.

Grodal S, Gotsopoulos A, Suarez F F. The coevolution of technologies and categories during industry emergence[J]. Academy of Management Review, 2015, 40(3): 423-445.

Delmestri G, Greenwood R. From a Cinderella into a Queen’: Radical status recategorization[J]. Administrative science quarterly, 2016, 61(4): 507-550.

Reuber A R, Dimitratos P, Kuivalainen O. Beyond categorization: New directions for theory development about entrepreneurial internationalization[J]. Journal of International Business Studies, 2017, 48: 411-422.

Autio E. Orchestrating ecosystems: a multi-layered framework[J]. Innovation, 2022, 24(1): 96-109.
Audio E. Orchestrating ecosystems: a multi-layered framework[J]. Innovation, 2022, 24(1): 96-109.

Shipilov A, Gawer A. Integrating research on interorganizational networks and ecosystems[J]. Academy of management annals, 2020, 14(1): 92-121.

韩进,李平,周海波.企业管理情境下生态系统理论框架与未来研究方向[J].管理学报,2022,19(01):139-149.
Han Jin, Li Ping, Zhou Haibo. Framework and future research directions of ecosystem theory in the context of enterprise management[J]. Journal of Management, 2022, 19(01): 139-149.

黄欣卓, 米加宁, 章昌平, 等.科学数据复用研究的演化、知识体系与方法工具——兼论第四科研范式的影响[J].科研管理,2022,43(08):100-108.
Huang Xinzhuo, Mi Jining, Zhang Changping, et al. Evolution, knowledge system and method tools of scientific data reuse research - also on the impact of the fourth scientific research paradigm[J]. Science Research Management, 2022, 43(08): 100-108.

张翱, 孙久文. 数据流动、数据要素化与数字经济运行[J].科学学研究, 20250218.002.
Zhang Aosun, Jiuwen, Data flow, data factorization, and the operation of the digital economy [J]. Science Research 20250218.002.

Suarez F F, Grodal S, Gotsopoulos A. Perfect timing? Dominant category, dominant design, and the window of opportunity for firm entry[J]. Strategic management journal, 2015, 36(3): 437-448.
Suarez F F, Grodal S, Gotsopoulos A. Perfect timing? Dominant category, dominant design, and the window of opportunity for firm entry [J]. Strategic Management Journal, 2015, 36(3): 437-448.

Adner R. Ecosystem as structure: An actionable construct for strategy[J]. Journal of management, 2017, 43(1): 39-58.
Adner R. Ecosystem as structure: An actionable construct for strategy [J]. Journal of Management, 2017, 43(1): 39-58.

Teece D J. Business models and dynamic capabilities[J]. Long range planning, 2018, 51(1): 40-49.

焦豪.数字平台生态观：数字经济时代的管理理论新视角[J].中国工业经济,2023,(07):122-141.
The Digital Platform Ecosystem Perspective of Jia Hao: A New Perspective on Management Theory in the Digital Economy Era[J]. China Industrial Economics, 2023, (07): 122-141.

Mann G, Karanasios S, Breidbach C F. Orchestrating the digital transformation of a business ecosystem[J]. The Journal of Strategic Information Systems, 2022, 31(3): 101733.

Cusumano M A, Gawer A, Yoffie D B. The business of platforms: Strategy in the age of digital competition, innovation, and power[M]. New York: Harper Business, 2019.

葛安茹, 唐方成. 基于平台包络视角的平台生态系统竞争优势构建路径研究[J]. 科技进步与对策, 2021, 38(16): 84-90.
Ge Anru, Tang Fangcheng. Research on the Construction Path of Competitive Advantage in Platform Ecosystem Based on the Perspective of Platform Bundle Envelope[J]. Science & Technology Progress and Policy, 2021, 38(16): 84-90.