Inside TikTok: Culture, Strategy, Monetization, and More | Ray Cao
- Aisha Washington
- 1 day ago
- 7 min read
TikTok’s rise cannot be explained by short videos alone. Behind the app is an operating system built around rapid experimentation, global product development, algorithmic discovery, and close coordination among product, engineering, sales, and data teams. In this conversation, Ray Cao offers an insider’s perspective on how those elements work together.
Speaking with Lenny Rachitsky, Cao explores TikTok’s culture, hiring standards, planning process, localization strategy, creator economy, and advertising business. His account also reveals the tensions inside a fast-growing organization: autonomy requires discipline, speed can undermine hiring quality, and global scale depends on understanding local behavior in unusually fine detail.
From Google to TikTok: A Different Operating Rhythm
Cao describes TikTok as retaining a startup mentality even as it expands internationally. Employees are expected to keep learning, move quickly, and treat growth as an ongoing process rather than evidence that the company has arrived.
One principle he highlights is “context, not control.” Leaders should give people enough information to understand the problem, constraints, and desired outcome without prescribing every action. In theory, that allows employees to take ownership and make faster decisions. In practice, it works only when teams share sufficient context and the organization hires people capable of exercising sound judgment.
Compared with his experience at Google, Cao found TikTok more willing to adjust products around the needs of several constituencies at once: viewers, creators, and advertisers. He portrays product development as flexible and experimental, with features refined through market feedback rather than rolled out according to a single US-centered sequence.
That global orientation is central to TikTok’s strategy. Instead of assuming that success in one major market can simply be exported, the company deploys people and resources where local knowledge is needed.
Why Global Algorithms Still Need Local Expertise
TikTok’s recommendation system learns from behavioral signals, but Cao argues that machine learning alone does not explain the platform’s international performance. Local teams help identify which behaviors matter, which content categories deserve early attention, and how cultural preferences should shape product decisions.
The task is not merely translating an interface. It involves understanding what people in a particular market enjoy, how they discover products, and which creative conventions feel familiar. Cao gives the example of food content attracting particular interest in Japan, while consumer technology and electronics can resonate strongly in Japan and parts of Southeast Asia. In the United States, TikTok’s early association with lip-syncing broadened over time into shopping, recommendations, and product discovery.
These differences affect how the recommendation engine is initially guided. A market may need the right creators, topics, and use cases before the system has enough meaningful activity to learn from. Cao’s broader point is that globalization requires deliberate seeding and continuous interpretation, not passive deployment.
Human judgment therefore remains important. Data can expose a pattern, but experienced operators still have to determine whether it reflects a durable shift, a short-lived trend, or a misleading signal. TikTok’s approach, as Cao presents it, combines machine learning with local market expertise and commercial judgment.
Autonomy Depends on Curiosity and Discipline
TikTok’s principles encourage employees to cross organizational boundaries, challenge established assumptions, and act without waiting for detailed instructions. Cao says this can reduce silos and make collaboration more fluid, especially when everyone understands the same goals.
Yet autonomy does not mean a lack of structure. TikTok uses objectives and key results, or OKRs, to align teams and connect their work to company priorities. Cao emphasizes the continuing need to improve both the quality of those objectives and the relationship between inputs and outcomes.
The hiring profile follows the same logic. Cao values curiosity, adaptability, and a genuine desire to learn, but he places equal weight on discipline, prioritization, and the ability to collaborate. A person who generates ideas without organizing their work may struggle just as much as someone who reliably executes but cannot adjust to change.
TikTok’s “always day one” mentality reinforces this expectation. Employees are encouraged to behave as though the company is still earning its future rather than defending an established position. Cao acknowledges that this environment can require a lifestyle adjustment and may not suit everyone.
How Cross-Functional Work Happens
Cao describes close cooperation among engineering, product management, sales, and data science as a competitive advantage. TikTok uses structured “doc reading” meetings in which participants review written material together before discussing decisions. The format gives attendees a shared factual base and reduces the risk that a meeting will be dominated by whoever speaks first or holds the most senior title.
Product managers and research-and-development teams are also encouraged to meet advertisers, clients, and partners directly. This exposure helps technical teams see the practical difficulties customers face instead of receiving those problems through several layers of internal interpretation.
TikTok did not invent every element of this system. Cao says the company has learned selectively from other organizations, including Amazon’s “day one” philosophy and memo practices, as well as Google’s use of OKRs. The aim is not exact imitation. TikTok adapts outside practices to its own pace, structure, and international footprint.
Organizing Product Teams for Global Growth
TikTok’s product organization was designed to recruit international talent and remain close to important markets. Cao notes that substantial engineering and product resources were based on the US West Coast, including Los Angeles and San Jose, while many go-to-market leaders worked from New York. Singapore became another major center supporting Southeast Asia.
The arrangement was not treated as permanent. As markets grew, TikTok repeatedly adjusted the relationship among product, engineering, and commercial teams. In some situations, product specialists were embedded more closely with sellers so they could develop a stronger understanding of customer demand.
Cao recognizes the value of organizational stability, but he argues that a fast-growing company cannot preserve team boundaries at the expense of results. Structure should clarify responsibility while remaining responsive to changes in the market.
Lessons From Scaling the Go-to-Market Organization
One of Cao’s most instructive stories concerns an effort to hire 100 people within six months. The aggressive target created pressure to compromise on candidate quality. Instead of accelerating expansion, unsuitable hires introduced friction and slowed the organization down.
A second problem was insufficient market context. A go-to-market plan may look coherent internally while failing to reflect the realities facing sales teams and customers. Cao learned that strategy cannot be separated from the conditions in which it must operate.
The third lesson concerned leadership distance. As teams expand, leaders can become consumed by management and lose contact with customers. Cao argues that leaders must continue speaking with clients and examining market details. A manager who only manages people may gradually lose the practical edge needed to recognize important problems.
Together, these mistakes underline a recurring theme: speed is valuable only when paired with high standards, current information, and direct exposure to reality.
Hard Work, Career Growth, and Personal Choice
Cao speaks openly in favor of sustained effort and, at times, long hours in a startup environment. He sees intense work as a possible source of meaningful achievements and career-defining experiences.
At the same time, he frames that commitment as an individual decision rather than something a company should simply impose. The distinction matters because a demanding culture can create opportunity for some employees while becoming unsustainable for others.
His preferred measure of a healthy organization is not permanent retention. Cao wants employees to be happy, develop new capabilities, and advance their careers—even if that growth eventually takes them elsewhere.
TikTok’s Creator Economy and Monetization Model
TikTok’s business depends on a functioning relationship among audiences, creators, and advertisers. Creators can earn through several channels, including platform reward programs, sponsored partnerships with brands, and virtual gifts received during livestreams.
For businesses, this creator ecosystem offers more than conventional ad inventory. Working with creators can give a brand access to established communities, platform-native storytelling, and credibility that polished corporate content may lack.
The platform also supports both broad awareness and measurable action. Cao describes TikTok as aspiring to become a full-funnel advertising platform: one that can introduce a product, stimulate interest, and contribute to purchases or other conversions. The popularity of “TikTok made me buy it” illustrates how entertainment, recommendation, and commerce can converge in the discovery feed.
How Creators and Brands Can Succeed on TikTok
Cao’s advice for creators begins with authenticity. Rather than importing a carefully managed persona from another network, people should identify something distinctive they can demonstrate repeatedly. A familiar song, for example, can become newly engaging when a creator changes the lyrics to reflect personal experiences or a particular community.
Success also requires participation in TikTok’s culture. Sounds, visual effects, challenges, and fast-moving formats provide useful creative ingredients, but copying a trend mechanically is rarely enough. Creators need to interpret it through a recognizable skill, perspective, or identity.
For advertisers, Cao recommends building a creative capability specifically for TikTok. Content made for Instagram or another social network may not translate because TikTok distributes videos through discovery rather than primarily through existing social connections.
A serious test should include:
A business account with regular organic publishing
Multiple creative concepts rather than minor edits of one advertisement
Roughly 10 new videos per week during an initial learning period
At least a month of testing before drawing firm conclusions
Frequent use of reporting data to refine creative and campaign structure
He also advises against beginning with narrow audiences or an excessive emphasis on remarketing. Broader targeting can give the recommendation system more room to identify responsive viewers and help advertisers learn who is interested in the product.
The essential method is “test and learn.” TikTok’s trends move quickly, so advertisers must be prepared to produce, evaluate, and replace creative work at a faster cadence. Tools such as CapCut can lower production barriers, but volume alone is insufficient. Each experiment should improve the team’s understanding of the audience, message, or format.
The Larger Lesson From TikTok
Cao’s account presents TikTok as a company built on productive combinations: algorithms and human judgment, global infrastructure and local expertise, autonomy and discipline, annual planning and continuous adaptation.
Its practices are not automatically transferable to every organization. The pace can be demanding, repeated restructuring carries costs, and experimentation can become wasteful without clear learning goals. Still, TikTok offers a compelling example of how a product company can connect technical systems with market knowledge.
For leaders, creators, and advertisers, the most useful takeaway is not simply to move faster. It is to shorten the distance between action and evidence—while preserving the judgment needed to understand what the evidence means.
