📈 economics

Research on the Entrepreneurship and Transmission Evolution Mechanism of Platform Enterprises Based on Complex Networks

This study utilizes complex network modeling to demonstrate that while hub-and-spoke structures maximize entrepreneurial participation, market-oriented networks ensure the fastest transmission stability, with both organizational learning and platform-initiated entrepreneurship serving as critical drivers of structural evolution in platform enterprises.

Original authors: Chun Huang, Rui Feng, Yan Cai, Kai Sun

Published 2026-07-01

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Chun Huang, Rui Feng, Yan Cai, Kai Sun

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a giant, bustling digital marketplace where companies don't just sell products; they sell the idea of starting new businesses. This paper is like a computer simulation lab where researchers built three different "worlds" to see how the spark of entrepreneurship spreads from one company to another.

Here is the story of what they found, explained simply:

The Three Different Worlds

The researchers created three distinct types of networks (or "worlds") to see how ideas travel:

The "Star" World (Hub-and-Spoke): Imagine a massive, super-popular influencer (the Platform) in the center. Hundreds of smaller companies are connected only to this influencer, not to each other. It's like a starfish: one big center with many arms.
The "Marketplace" World (Market-Oriented): Imagine a giant flea market where everyone is roughly the same size. Companies bump into each other randomly, trading ideas and resources. No single boss runs the show; it's a free-for-all.
The "Hybrid" World: This is a mix. There is a big boss, but there are also smaller, semi-independent teams working under them. It's like a large corporation with several distinct, semi-autonomous startups inside it.

The Experiment: Lighting the Fuse

The researchers wanted to see how "entrepreneurship" (the act of starting new ventures) spreads through these worlds. They used a computer program to simulate two main scenarios:

Who starts the fire? Does the big "Star" influencer start the trend, or do random small companies start it?
How many start at once? If 5% of the companies start a new business, does it spread differently than if 15% start at the same time?

The Big Findings

1. The "Star" World creates the most new businesses.
When the big central platform started the trend, the "Star" world ended up with the highest number of new companies. Because everyone is connected to the powerful center, it's easy for them to copy the big boss. It's like a celebrity starting a dance trend; everyone in their fan club copies it immediately.

2. The "Marketplace" World spreads the news the fastest.
Even though the "Star" world created more businesses, the "Marketplace" world got there quicker. In the random flea market, ideas hop from person to person in many different directions at once. It's like a rumor in a crowded room; it travels fast because everyone is talking to everyone else, even if the rumor doesn't reach as many people as the celebrity's post.

3. The "Hybrid" World is in the middle.
It wasn't as fast as the Marketplace, but it didn't create as many businesses as the Star world. It was a balanced, middle-ground approach.

4. More starters = Faster results.
If you start with more companies already doing the new thing (increasing the initial number from 5% to 15%), the whole network catches on faster. It's like lighting a campfire: if you have more kindling to start with, the fire spreads to the whole forest much quicker.

5. Who starts it matters (at first).
If the big Platform starts the trend, it spreads very fast at the beginning because everyone is watching the boss. If a random small company starts it, it takes a little longer to get going because they don't have as many eyes on them. But eventually, once the trend is hot, it doesn't matter who started it; the speed evens out.

The Bottom Line

The paper concludes that if you want to create the most new businesses, you need a strong central leader (a Hub-and-Spoke network). But if you want the idea to spread quickly across the board, a loose, random network (Market-Oriented) is the most efficient way to get the word out.

The researchers used a tool called "NetLogo" (a simulation software) to run these digital experiments, proving that the shape of the network determines how fast and how far the spark of innovation travels.

Technical Summary: Entrepreneurship and Transmission Evolution Mechanism of Platform Enterprises Based on Complex Networks

Problem Statement
The study addresses the challenge of understanding how platform enterprises foster corporate entrepreneurship and how this entrepreneurial behavior transmits and evolves within industrial clusters. While existing literature examines platform entrepreneurship from the perspectives of core firms or general network relationships, there is a lack of systematic research utilizing complex network theory to model the structural evolution of these transmission mechanisms. Current social network analysis methods are insufficient for quantitatively measuring the internal complex relationships and nonlinear dynamics inherent in entrepreneurial networks. This paper seeks to "uncover the black box" of platform enterprise entrepreneurship by simulating how different network structures influence the speed, stability, and scale of entrepreneurial transmission.

Methodology
The research employs a multi-agent simulation approach using NetLogo 6.1.0 to model the dynamic evolution of entrepreneurial networks. The methodology integrates time series analysis with three distinct complex network models, each mapped to a specific type of platform enterprise network structure identified in the literature:

Market-Oriented Networks: Modeled using the Erdős-Rényi (ER) Random Graph model. This represents networks with homogeneous firms and randomized resource exchanges, characterized by uniform degree distribution and stochastic connections.
Hybrid Networks: Modeled using the Local-World Evolving (LN) network model. This reflects structures where firms operate within specific subnetworks or localized domains, selecting partners based on local availability and complementary skills, resulting in moderate heterogeneity.
Hub-and-Spoke Networks: Modeled using the Barabási-Albert (BA) Scale-Free network model. This represents a structure centered on a single dominant platform enterprise, characterized by power-law degree distribution, preferential attachment, and high dependency on a central hub.

Simulation Rules and Algorithms
The simulation incorporates two primary transmission algorithms:

Threshold Diffusion Model: Enterprises adopt entrepreneurial behaviors only when influenced by successful peers beyond a specific critical threshold. This accounts for organizational learning and the evaluation of native versus emergent thresholds.
Neighbor Effect Model: Specifically under an active influence paradigm, the model simulates how connected enterprises consciously disseminate innovations to neighbors, strengthening relationships and accelerating implementation.

The simulation tracks four key performance indicators to evaluate mechanism evolution:

Transmission Success Rate ( $S$ ): The number of enterprises successfully adopting innovations.
Transmission Stabilization Time ( $T$ ): The duration required for entrepreneurial activities to reach equilibrium.
Transmission Amplification Coefficient ( $E$ ): The cumulative number of entrepreneurial instances within the network.
Transmission Concentration ($CV$): The centralization tendency of entrepreneurial behaviors.

Experiments were conducted with up to 500 nodes (enterprises), varying the initial proportion of entrepreneurial firms (5%, 10%, 15%) and the source of transmission (platform enterprises vs. random enterprises).

Key Results
The simulation yielded the following findings regarding the evolution of entrepreneurial transmission:

Network Structure and Participation: Hub-and-spoke (BA) networks generate the highest number of entrepreneurial firms. Under identical initial conditions, hub-and-spoke networks outperform hybrid (LN) networks, which in turn outperform market-oriented (ER) networks in terms of the total volume of entrepreneurial activity.
Transmission Speed and Stability: Conversely, when controlling for the proportion of startups, market-oriented (ER) networks achieve transmission stability the fastest, followed by hybrid networks, and finally hub-and-spoke networks. Market-oriented networks exhibit the shortest stabilization time and the highest cumulative amplification coefficient, suggesting that their complex, multi-chain connections allow for faster signal propagation and equilibrium.
Influence of Spreading Sources: Using platform enterprises as the initial spreading source accelerates transmission in the early stages due to their core competitiveness and resource advantages, which reduce information asymmetry. However, in the mid-to-late stages, the spreading speed converges regardless of the source.
Initial Conditions: Increasing the initial proportion of entrepreneurial enterprises shortens the overall stabilization time and reduces the concentration of entrepreneurial behavior, indicating that a larger initial base creates a stronger entrepreneurial atmosphere and accelerates the diffusion process.

Significance and Claims
The paper claims to be innovative and pioneering in the empirical study of corporate entrepreneurship and its transmission effects by applying complex network theory to this domain. Its primary contributions include:

Theoretical Integration: It bridges the gap between entrepreneurial network theory and complex network dynamics, moving beyond static social network analysis to dynamic, quantitative simulation.
Structural Classification: It provides a systematic classification of entrepreneurial networks (Hub-and-spoke, Hybrid, Market-oriented) and maps them to specific complex network models (BA, LN, ER) to simulate their distinct evolutionary logics.
Mechanism Revelation: The study reveals the inherent logic of entrepreneurial evolution, demonstrating that while hub-and-spoke structures are superior for generating a high volume of entrepreneurial firms, market-oriented structures are more efficient for the speed and stability of transmission.

The authors conclude that both the size of the startup base and the initiation of entrepreneurship by the platform itself act as catalysts for transmission, collectively accelerating business development. The research offers a new perspective for understanding how platform enterprises can optimize their network structures to balance the goals of maximizing entrepreneurial participation and ensuring rapid, stable transmission of innovative behaviors.

The Three Different Worlds

The Experiment: Lighting the Fuse

The Big Findings

The Bottom Line

Technical Summary: Entrepreneurship and Transmission Evolution Mechanism of Platform Enterprises Based on Complex Networks

More like this