How Zhang Yiming’s Multi‑Thinking Model Fueled ByteDance’s Short‑Video Dominance

01 All in Short‑Video

In 2016, while Tencent Weishi and Weibo’s short‑video apps were gaining traction, Zhang Yiming convened senior colleagues for a third discussion on whether ByteDance should enter the short‑video market.

The product manager argued that short‑video no longer offered a first‑mover advantage and that ByteDance lacked specialized talent, making success unlikely.

In contrast, the product director highlighted a SWOT analysis: the market was still in its early stage, no single player dominated, and ByteDance’s powerful recommendation engine gave it a strong cold‑start advantage.

Ultimately, Zhang Yiming decided to go “All in” on short‑video, launching multiple products (Douyin, Huoshan, TikTok) domestically and abroad, and acquiring leading competitors.

Four years later, Douyin and Huoshan dominate China, TikTok leads globally, and ByteDance operates in over 150 countries and 75 languages.

02 The Multi‑Thinking Model – Zhang Yiming

Many tech leaders, such as Steve Jobs, Elon Musk, Lei Jun, and Zhang Yiming, exemplify “multi‑thinking” – using multiple mental models to solve problems.

Zhang’s early career as a crawler programmer at Kuxun showed his “no‑boundary thinking”: he helped colleagues beyond his assigned tasks, shared code knowledge, and quickly rose to lead larger backend teams.

This mindset led to his first investment from Wang Qiong and later to the founding of ByteDance.

His “optimal‑solution thinking” treats resources as quantifiable, selecting the best option based on defined dimensions, as illustrated by his college choice and his refusal to sell Toutiao to Tencent.

The “super‑allocation thinking” reflects concentrating strategic resources to achieve breakthroughs, evident in ByteDance’s aggressive product launches, acquisitions, and content‑creation initiatives.

03 Why Use Multi‑Thinking?

Relying on a single perspective is like blind men feeling an elephant; it yields a limited view. Charlie Munger warned that a single tool (a hammer) makes everything look like a nail, but without a toolbox (multiple models), one cannot address diverse problems.

Prominent figures such as Jeff Bezos, Elon Musk, Warren Buffett, Cheng Wei, and Wan Weigang champion multi‑thinking.

04 Becoming a Multi‑Thinking Expert

Effective use follows a cycle: learn models → analyze events → apply appropriate model → verify → adjust model → produce conclusions.

Practitioners must understand each model’s scope, select the right one, test conclusions, and iterate.

05 Core Thinking Models Across Disciplines

Examples include natural selection (biology), the invisible hand (economics), first‑principles (philosophy), and entropy (thermodynamics). A full list of 61 models from 22 fields is available via the “Technology Leadership” channel.

In summary, mastering multiple thinking models enables leaders to swiftly perceive the essence of complex issues and make superior strategic decisions.