Insta360 Rings the Bell: A 10-Year Retrospective with 刘靖康
Summary
Insta360’s listing changed its downside risk, not its appetite for innovation. After listing on the STAR Market on June 11, 2025, the company raised RMB1.938B; its R&D expense ratio has stayed above 13% for years, but 刘靖康 is no longer willing to go all in on a bet where getting it wrong means liquidation. High-risk projects remain on the table, but business design, cash reserves and tighter planning must protect investors. His message to public-market investors is unusually counterintuitive: “Be more optimistic about us long term, and more objective about us short term”—and do not get harvested by supplier-concept speculation.
Faced with DJI’s price war, Insta360 chose not to match prices, forcing the product to answer why customers would buy a smaller brand at a higher price. About 3 months after Insta360 launched its action camera at $450, its rival cut a $400 product to $300; the gap at Chinese retail stores reached RMB700-800. 刘靖康 believes matching the cut would only invite a better-funded giant to cut again, so he reframed the fight as a product question: “We can only force ourselves to make this thing extraordinarily good.”
刘靖康 does not equate short-term scale with superior capabilities; what he really wants is “not to race for the first flow, but to keep the river flowing.” At the time of recording, Insta360 had roughly RMB4B in annual revenue, versus DJI’s roughly RMB50B of revenue and RMB8B of profit. Though Insta360 says it has surpassed GoPro in revenue, he does not believe its underlying capabilities have done so. DJI is both “the greatest misfortune” and “the greatest luck”: a powerful rival exposes gaps in supply chain, cost, organization and talent, while offering a capability-building path to study without copying.
Insta360 is moving from a founder-led super-product-manager company toward an organization that scales through generals, systems and a talent supply chain. Headcount rose from roughly 600 three years ago to about 2,000 in 2024 and more than 3,000 in 2025, spanning over 600 job types. The company once prioritized engineers, process and checkpoints, only later recognizing that product managers and systems engineers could jointly determine a product’s ceiling and floor. 刘靖康’s order is “employees first, customers second, shareholders third,” tested through context not control, rotations, virtual product lines and, when conditions allow, orderly spinouts for exceptional talent.
The Ace Pro 2 fogging recall was the most nerve-racking battle of the past decade—and a test of whether the organization could save itself after separating from its founder. The flagship, developed with tens of millions of RMB in investment, was halted and recalled immediately after launch because it fogged up in some scenarios. The team apologized, replaced units and issued refunds before using customer research to launch a street-photography kit; by “618,” demand exceeded supply. 刘靖康 accepts that he managed quality outcomes but not preventive processes. What exhilarated him was that he had almost no involvement in the successful solution: “This company may be able to make money even without me.”
AI matters to Insta360 not just for automatic editing, but for rewriting the camera’s human-machine interaction from capture and scene understanding through the final cut. Its automatic-editing capability has advanced from roughly 60 points a few years ago to “not quite 70,” with a target above 80; 刘靖康 spends roughly 10%-15% of his time on the intersection of AI and products. Agents and natural language could resolve the trade-off between simpler interfaces and fewer functions, while proprietary hardware and cloud services let Insta360 tackle automatic shooting, scene understanding and more complex cloud inference at once. He describes large models as “everyone suddenly having a Gatling gun dropped into their hands.”
Insta360 ultimately wants to create new categories, not become a smaller DJI or win share through discounting. 刘靖康 sees imaging as an “open world” like a game: creating new experiences comes before efficiency, while original value funds R&D, brand building and the next generation of products through gross margin. The roadmap already contains 3 fully original, world-class targets. Going public and carrying obligations to employees, customers and investors weakens the courage of having “nothing to lose,” but his hedge is “you only live once”: he fears spending a lifetime following other people’s ideas more than he fears having less cash.
Deep dive
1. The Listing Was a Milestone in a Ten-Year Voyage, Not an Emotional Climax
卫诗婕 juxtaposed the 2 recordings: the first, in August 2024, when Insta360 was doubling year after year and confronting DJI head-on; the second, around its June 2025 listing, observing how 刘靖康’s mindset had changed as he faced world-class competition.
On June 11, 2025, Insta360, founded 10 years earlier, listed on the STAR Market. By the program’s opening description, the company had become the global No. 1 in 360-degree cameras, No. 2 in action cameras, and ranked No. 3 in 2025 A-share fundraising at that point.
Yet 刘靖康 could not summon the excitement of finally dating one’s ideal partner or getting into one’s dream school: “It felt like completing a milestone.” An A-share listing may happen only once in a lifetime. He explored ringing the bell in a T-shirt, but ultimately wore a suit.
On being called “the youngest chairman in STAR Market history,” he insisted on adding “one of.” He was only 29 when he appeared before the Shanghai Stock Exchange 4 years earlier, but still believes the innovation board will eventually see someone younger.
2. 刘靖康 Wants to Preserve Not Caprice, but Informed Independence
Between the 2 interviews, he believes his tactical understanding has changed, while his underlying character has not. What worries him more is that the environment and competitors might force him into becoming someone he does not want to be.
He uses the Japanese word “わがまま” to describe the ideal state: take in everyone’s information, then decide according to one’s own risk tolerance and interests. That means neither shutting out opinions and acting blindly on one’s own, nor becoming “the executor of the aggregate vote.”
卫诗婕 summarized it as a refusal to become conventional but correct: he wants to preserve non-consensus judgments when fully informed. 刘靖康 agreed. He does not want his independent understanding of a matter or the future overwritten by democracy or composite opinion.
After the listing, he still wants to remain “abstract,” provided he stays within the law and the rules. Being accountable to investors does not require official-sounding communication; if his actions genuinely benefit shareholders, employees and customers, he sees no need to adopt a different personality.
3. A Decade in Shenzhen Turned “Making Tokyo Your Place” into a Living Metaphor
When Insta360 put up the sign on its building, 刘靖康 quoted the Japanese drama Tokyo Tower: “Kid, if you’re going to Tokyo, then make Tokyo your place.” The line summed up his 10 years in Shenzhen and repeatedly encouraged him during low points.
After graduating from university, he and the team first started a company in Nanjing, then moved to Shenzhen for its supply chain and talent pool. At the time, the move was neither a dream of settling down nor a source of fear—more a fresh, risky business choice.
The protagonist of Tokyo Tower draws erotic manga to survive, mirroring the forced choices that follow entrepreneurship: as a student, you can freely do what you like; once in society, there are always tasks “you do not necessarily like.”
His motivation does not primarily come from family, but from the intersection of 3 things: doing what he likes, creating social value and making a lot of money. After 2019, “making a lot of money” fell in priority; surviving economic cycles, changing the world and “building the company as a product” gradually became stronger motivations.
4. Making Money Is an Operating Test, Not a Founder Wealth Target
刘靖康 believes there are many ways for an individual to become rich, including by sacrificing a company’s long-term interests. A company has to make money because profit validates the product, market selection and organizational efficiency, while funding the next phase of investment.
To support the company, he once borrowed from a bank in his own name at annual interest in the 7-figure range, with most of his monthly salary going toward interest payments. Becoming a father naturally made him hesitate, but from the founder’s seat, not stepping in would have meant putting his personal interests above the collective interest.
At the same time, he refuses to dress up the decision as selflessness: “The core point is that the company’s fundamentals are actually quite good.” The resulting personal cash-flow strain was real, but compared with the company’s development, it “didn’t amount to anything.”
5. Behind the Word “Calm” Was an Organizational Self-Audit Using Data
For his birthday, the head of BI turned 刘靖康’s Feishu records into a “JK scolding analysis and trend report.” Negative emotion was divided into 4 levels; peaks usually came in the month before a new product launch, with the densest activity in product-acceptance groups.
The report also summarized his anger patterns, showing him that emotion often obscured the point itself. On one call, he lost his temper in front of everyone, prompting a colleague to resign on the spot. That became the direct reason the office Hello Kitty board was labeled “Calm.”
刘靖康 admits Insta360 is still in its “acne years,” with even more things he does not want to do than 2 years ago. But he does not want to hide his rough edges behind an entrepreneur’s protective coloration; staying authentic internally and externally is itself a choice.
The CEO reminds him to speak less recklessly and manage his health and time. He jokes back that if he keeps everything bottled up and “gets cancer from holding it in, that’s on you.” Underneath the banter, both agree that a stable founder is what keeps the company from losing speed.
6. Wearing DJI’s Shirt and Carrying a GoPro Bag Was His Rejection of Binary Competition
刘靖康 wore a DJI T-shirt during the recording and explained it bluntly: “Conflict and contradiction are the easiest things to spread.” But it also reflected his mindset: competition is not the whole job, and a rival does not have to be hated.
He has carried a GoPro bag for years. When the fight with DJI left him frustrated, he would watch GoPro’s fifth- and sixth-generation promotional videos late at night to “get himself fired up”; the sixth generation stirred him most.
He has also written “I love GoPro” and “I love Nick Woodman” on Twitter. A rival can compete for market share while inspiring him through its products, brand and founder’s spirit. The 2 things are not in conflict.
7. He Refused to Follow Price Cuts Because a Better-Funded Rival Can Always Reopen the Gap
After Insta360 launched a gimbal camera priced above RMB900, DJI cut its comparable product to around RMB800. About 3 months after Insta360 launched its action camera at $450, DJI cut a $400 product directly to $300.
The core message to consumers at Chinese stores was clear: Insta360 makes good products, but DJI is RMB700-800 cheaper. With a larger brand selling at a lower price, price had become part of the purchase decision.
刘靖康 refused to match the cut: once Insta360 follows, the rival only needs to cut again and the gap it meant to eliminate instantly returns. “Their pockets are deep. In principle, they can do this indefinitely. So what exactly are you trying to achieve?”
After refusing to cut prices, the question became “why would customers buy you?” Insta360 only needed to make its product, marketing and service distinctive enough for users to understand deeply what problem it solved. Pressure from the giant instead forced it to pursue something extraordinary.
8. Sales Were on Plan, but the Real Unclosed Gap Was Still GoPro
When Ace Pro went head-to-head with Action 4, sales met the original expectations but did not significantly exceed them. The review concluded that the issue was not simply pressure from DJI’s pricing; Insta360’s gaps with GoPro in product, brand, marketing, service and distribution had not been closed.
刘靖康 admits that he focused too heavily on DJI over the prior year. One reason was geographical: the 2 companies are only about 10 kilometers apart, while GoPro is 10,000 miles away. Another was that DJI was more proactive, continuously using new moves to pull Insta360’s attention.
卫诗婕’s scale comparison at the time was roughly RMB50B in annual revenue and RMB8B in profit for DJI, versus about RMB4B in annual revenue for Insta360. Looking back at Insta360’s entry into action cameras in 2017, 刘靖康 notes that GoPro’s revenue was once about 40x larger and still more than 10x larger in 2019, while stressing that the 2 competitive situations cannot be compared mechanically.
9. A Strong Rival’s Value Is Exposing Weaknesses, Not Creating a Tragic Founder Narrative
刘靖康 defines the company’s goal as “not to race for the first flow, but to keep the river flowing.” Short-term profit, revenue and scale can swing sharply; the companies remembered in the end are generally those that survive long enough and make it through cycles.
A DJI-like competitor sharply increases pressure while exposing weaknesses in R&D, supply chain, marketing and organization. The hardest part is not recognizing that weaknesses exist, but deciding amid the anxiety which one to fix first.
Startups survive day one on their strengths, but as the business expands they enter other people’s territory or attract attention to their own niche. Competition gradually becomes a question of “who has fewer weaknesses.” 刘靖康 regards a strong rival as a demanding teacher: the process hurts more, but long term it can train a more complete player.
He calls profit the funding source for capability building and the report card for operating efficiency, not the objective itself. Insta360 wants to maximize accumulated underlying capability rather than short-term profit.
10. DJI Offered Not a Standard Answer, but an Underlying Mechanism to Reverse-Engineer
On cost reduction, Insta360 might once have started with supplier negotiations. Studying DJI showed 刘靖康 that design, yield, process and supplier co-development are the source of the savings; in some cases, the company should support or invest in a supplier’s cost-reduction R&D.
For key materials, DJI develops or customizes chips and platformizes components. Multiple products use the same sensor, allowing centralized purchasing volume to reduce the cost of each SKU. 刘靖康 calls these practices only “the tip of the iceberg.”
On product strategy, he observes that DJI prioritizes eliminating weaknesses over building one exceptionally long strength. That may sacrifice differentiation in the short term, but is better suited to complex hardware, where fatal defects cannot be tolerated.
“Not knowing what you don’t know” usually becomes visible only after the cost has been paid, by which point the company is already a step behind. Systematically studying a leader’s design mechanisms, systems and methods can reveal the underlying gap before product and market differences become visible.
11. “The Greatest Misfortune Is Also the Greatest Luck,” Because Commercial Success Is Not Capability Success
刘靖康 fully agrees that Insta360 benefited from early tailwinds: the 360-degree market was small, serious competitors were scarce, and the global economy, capital markets and business environment were less harsh than they later became. “Today’s greatest misfortune is having DJI in the category, but the greatest luck is also having DJI.”
In competing with GoPro, Insta360 largely had to build its own road and figure out products and tactics. DJI provided a road that had already been paved as a reference. Insta360 need not copy it, but can ask why each system was designed that way.
Even though Insta360’s revenue has surpassed GoPro’s, 刘靖康 does not believe it has fully overtaken GoPro in capability. Commercial results can come from market, category or policy tailwinds; they do not prove the underlying foundation is solid.
The next decade also adds AI as a huge variable. If the goal is certainty of outcomes, the outlook may not be more optimistic than in the past. If entrepreneurship is a growth game, strong rivals and technological change will make the next 10 years “much more interesting.”
12. A Generational Advantage Can Buy Patience, but Cannot Prove a Young Company Will Survive
刘靖康 recalls the logic behind IDG’s post-90s founder fund: it was not betting on a particular sector, but on the strongest players in each generation. Outcomes such as Bilibili and Qingyan made him believe that age cohorts can sometimes provide a different long-term perspective.
He uses 黄峥’s thinking on generational competition to argue that instead of watching the giants of the previous generation every day, a founder should make sure to be among the strongest of his own generation. He is 10 years younger than 汪滔, which means the long-term pressure is “not that great.”
卫诗婕 asks: millions of small players have been crushed by giants, so why should Insta360 be Pinduoduo? 刘靖康 admits that “looking around, there don’t seem to be any more capable people our age” is only half a joke; logically, it does not prove survival.
What the company can do is design a safety boundary: assume the rival takes every extreme action, and still stay at the table. Failure can make the company stronger, but only becoming a better version of itself while remaining in the game has meaning.
13. Challenging Giants Comes from Both Mission and a Once-in-a-Lifetime Opportunity
刘靖康 admits there is no way to prove in advance that the company will not die. He can only use business plans, reserves and mechanisms to reduce the probability that a temporary failure knocks Insta360 out. The inability to prove safety is not a reason to reject the challenge.
Insta360 went global from day one. A group of university graduates with no overseas study or startup experience turned international expansion into a success through innovation in product, marketing and service. He believes the company has shown the industry that “as long as you take the right path, this can still be done.”
The next thing he wants to prove is that even against DJI, Insta360 can become No. 1 in market share while leading in both product and price. If it succeeds, it would give more founders without prestigious backgrounds a repeatable source of encouragement.
Asked whether he fears a prolonged price war with a giant, he answered directly: “Of course I’m afraid, of course I’m afraid,” before adding, “汪滔 has never gone up against an opponent like me either.” His self-definition is not a more vicious rival, but “an opponent who does not hurry.”
14. New Businesses Must Create Both Commercial and Capability Returns
Insta360’s original criteria for a new category included gross margin of roughly 50% or more, annual sales potential in the RMB10B range, and a clear remaining customer pain point. Smartphone gimbals and video-conferencing cameras have reached the market’s top tier and turned profitable, covering their respective long-term investment, though organizational issues mean the commercial result has not been maximized.
刘靖康’s full definition of a good business is that customers have a need, the company can solve it with a reliable and efficient offering, the business generates enough cash to operate sustainably, and it accumulates transferable capabilities for the future.
He will not enter markets such as water or ordinary consumer goods simply because they make money. If demand is not deep, differentiation rests mainly on distribution, brand and supply chain; Insta360’s real strengths are technology, product and marketing, so expanding casually would scatter its capabilities.
A later entrant’s product must make the gap visible “at a glance.” Some projects could have launched 6 months or even 3 quarters earlier but were still sent back. If competitiveness was insufficient, he would rather kill the project even if it could sell at a low price.
15. A Project Costing Tens of Millions Can Still Be Killed Because Sunk Costs Cannot Hide Product Weaknesses
Insta360 once developed a modular handheld gimbal camera: remove one module and replace it with another, and it became a 360-degree camera. The project absorbed tens of millions of RMB. It had a competitive strength in imaging, but obvious weaknesses—including becoming hot in the hand—ultimately kept it from launching.
Scrap from the project continued to appear in quarterly operating reports for years, declining over time and reminding 刘靖康 of the decision. But he believed launching it would consume marketing, supply-chain and service resources, producing an even worse return.
Competing purely on value with lower margins would not fit the brand or help capabilities grow. The exception would be designing a $400 product to be a genuine $250 product at the same margin, because that would force supply-chain integration, platformization and design standardization.
16. GoPro Karma Proved That One Fatal Flaw Can Overwhelm Every Strength in Complex Hardware
Reviewing the drones and gimbal competitors DJI had defeated, 刘靖康 found many products with pronounced strengths and weaknesses. Capital pressure then forced single-business companies to rush toward launch, using new products to replenish cash flow.
He keeps GoPro Karma on his desk as a reminder. He estimates the project consumed $150M-$200M. Its modular design and integrated remote were innovative, but a vibration issue could cut power midair, causing the drone to fall and forcing a recall.
Drones are the opposite of “a flaw does not obscure the brilliance”: when safety is involved, one flaw can erase all the innovation. The key is not to become intoxicated by one’s own product, or let shareholders and schedules force the decision; the team must keep testing whether the product meets consumer expectations.
Part of what he respects about DJI is its patience. Many projects are rumored to have been killed, with the company willing to sacrifice short-term returns for long-term completeness. Insta360 also wants to become a company that keeps innovating in the category and inspires others.
17. The “Capital Prisoner’s Dilemma” Turns Consensus Categories into a Double Kill of Margin and Talent Costs
刘靖康 sees brutal competition as a necessary phase, but not a permanent condition. Capital naturally needs to appreciate and funds must invest; when investors fail to study demand beneath the surface, capital concentrates on the easiest consensus story to explain.
Large pools of money draw smart people into the same category, making products and services increasingly homogeneous. Homogeneity compresses gross margin, while further capital inflows raise talent costs, producing a double kill: lower gross margin and higher labor costs.
He calls it “capital’s prisoner’s dilemma”: every dollar wants to compound, yet a lack of training leads investors to make the same short-term choice. Only after investing heavily without returns might investors learn to search for non-consensus opportunities.
Founders invest not abstract capital, but time, resources, youth and teams. If they fail to choose a category carefully, they too exhaust their resources in homogeneous competition, until only a few remain at the table and the rest are washed away.
18. The Most Dangerous Consequence of Gray-Zone Competition Is Shifting Attention from Customers to Rivals
刘靖康 describes unhealthy competition as supply-chain and channel “bloc formation”—in effect, forcing a binary choice—as well as information warfare and corporate espionage. He declines to elaborate.
Insta360 has not abandoned its bottom line. Whenever a rival crosses it, his first response remains to inspect the company’s own weaknesses and repeatedly ask: would holding a given principle obstruct progress? Is it a shortcut that can be abandoned, or a narrow bridge that must be crossed?
He sees caring for partners and attacking enemies as 2 sides of the same coin, not contradictions. The deeper problem is having attention captured by the rival. Seeing a competitor’s good news can trigger envy; if emotion is driven by competition for long enough, a person can gradually stop being himself.
王兴’s metaphor therefore resonates: “Your competitor is your rear-view mirror. You need to glance at it from time to time, but you can’t drive while staring at it.” Competitive information must be watched, but customers and the road ahead should determine direction.
19. Business Ethics Are Self-Imposed Rules, but Long-Term Failure Usually Still Comes from Failing to Do Your Own Job
刘靖康 has gradually learned “not to take one’s own experience too seriously.” Looking across competitive history, he has not seen many companies truly destroyed by underhanded tactics. Most failures came neither from an overly strong rival nor from gray-zone behavior itself, but from failing to execute key work internally.
He used to focus more on the second half of “business ethics.” Today he treats ethics as part of the rules of business. Faced with espionage, he may have to set aside empathy and become tougher; faced with comparative marketing, he may have to explain where others fall short.
The latter is not necessarily unethical, but it sacrifices the pride of staying above the fray. Talking only about one’s own strengths is more dignified, but when a rival is already making comparisons, silence leaves consumers unable to judge and forces Insta360 to follow the rules of reality.
What he really wants is a 100-meter track: everyone focuses on the finish line and their own speed, without tugging at the runner in the next lane. Demand can continue to be developed, and rivals can inspire one another—“use those moving in the same direction as a mirror, and those moving toward you as the truth.”
20. Insta360’s Core Advantage Is Understanding Demand Earlier, Then Engineering the Time Lead
刘靖康 summarizes the competitive advantage as understanding earlier than rivals what customers need now or will need in the future. Understanding alone is not enough; software, hardware and self-developed or customized chips must turn that insight into a deliverable time lead.
Technology planning must balance performance, cost and reliability while maintaining sufficient reserves. The company can absorb a wrong call in one direction, but a broad planning error can cause it to miss the entire product window.
Early 360-degree cameras forced Insta360 to accumulate extensive image-enhancement algorithms and engineering capabilities, which later supported Ace Pro. By his account during the recording, its night image quality was better than DJI Action 4 and GoPro HERO12 and HERO13, creating a “discontinuous advantage.”
Consumer electronics rarely have the two-sided markets and network effects seen on the internet. He cites Apple’s developer ecosystem, software experience and high-paying users as an exception. Insta360 is exploring longer-cycle moats, but he is unwilling to reveal its real “trump card” to competitors.
21. The Software Moat Is Moving from Code to Data, Training and Inference Efficiency
刘靖康 acknowledges that pure software advantages are quickly caught, but iteration speed is still constrained by organizational collaboration, the accuracy of demand discovery and software architecture—not simply by hiring a few more senior engineers.
In the past, software advantages often came from humans writing better algorithms and rules. New software is increasingly trained on data, requiring high-quality data, effective training methods and models suited to the task.
Models must also infer efficiently on-device or in the cloud at sufficiently low cost. Accuracy, performance and product experience are all necessary. Software competition has therefore expanded from code depth to data and infrastructure.
He believes the incremental opportunity from large models temporarily outweighs the threat. In the past, Insta360 had a rifle and others had pistols; now “everyone has had a Gatling gun dropped into their hands,” and whoever learns to use it first may reopen a discontinuous lead.
22. Going from 600 to More Than 3,000 People Is Like Changing Tires on a Highway
At the 2024 recording, Insta360 had around 2,000 people, versus roughly 600 3 years earlier. By the 2025 interview, it had more than 3,000 people across over 600 job types. The business needed to keep hiring, but the organization was not necessarily ready to absorb, train and give stage time to newcomers.
刘靖康 describes the state as “changing tires while driving”: the business cannot stop, but training, processes, management and the talent pipeline must all be upgraded at the same time. Organizational gravity has become tangible.
Facing giants, employee confidence cannot be built through slogans. During each new-hire training cohort, he and 李勇 explain why the company chose these businesses, its track record and why success is logically possible.
Theory and historical cases prove only that success is possible. Real confidence comes from positive feedback in battle: customer complaints are gradually resolved, small advantages keep appearing and the team wins incremental victories with its own hands.
23. Product Managers and Systems Engineers May Set Both the Ceiling and the Floor of a Hardware Product
Insta360 once prioritized increasing the density of frontline engineering talent, relying on departmental reviews, checkpoints and senior employees mentoring new hires. Later reviews found that it had underinvested in selecting and developing “generals.”
In consumer electronics, core product managers and systems engineers must both identify real demand and understand technical and system-wide feasibility. A wrong judgment does not merely lower the ceiling; an individual’s capabilities can themselves become a source of lost points.
This capability is difficult to replace with processes or reviews. 刘靖康 was not taught product judgment on day one; he developed it by repeatedly making mistakes in real projects. Candidates therefore need to be screened, but also given a stage with genuine responsibility.
The long-term solution is to develop second- and third-generation product managers while gradually building collective decision-making and succession mechanisms. But he acknowledges that every generation “has its own strengths,” and the next cannot inherit the previous generation’s wisdom intact.
24. Context Not Control Moves the Company from Managing Process to Picking the Right People and Giving Them Context
刘靖康 summarizes the pattern as “A players like to hire A players, while B players like to hire C players.” An alternative way to scale is to first select the best people, review the business plan, product definition, budget and milestones with them, then authorize them to build their own teams.
This resembles internal entrepreneurship or angel investing. The company provides resources, context and boundaries; the person in charge assumes first-seat responsibility. Inspired by ByteDance’s “context not control,” Insta360 began shifting from frontline control to general-led execution.
He makes clear that the principle does not require complete information transparency. Compared with issuing direct orders, managers should first provide decision-making context. The Business Plan system, for example, should bring people with a deeper understanding of frontline details into the design process.
When business participants jointly create the system, the result is usually more grounded than a top-down snap decision and more likely to solve practical problems. Delegation does not mean abandoning management; it shifts control from specific actions to people, context, budget and milestones.
25. The Biggest Mistake in Studying Huawei Is Attributing a Complete Capability Set to a Process Skeleton
刘靖康 believes companies approaching Huawei’s methods for the first time are easily dazzled by its enormous success and then persuaded by consultants that the result comes from a particular tool. They overestimate the method and expect short-term returns.
The learning cost is not the RMB millions or RMB10M consulting fee, but extensive employee training and the disruption of the entire organization. When immediate results fail to appear, a company with excessive urgency can easily abandon the effort halfway through.
Huawei’s methods emphasize process, technology reserves and customer value, which can conflict with short-term profit and efficiency. IPD is only the skeleton; module chiefs, LPDTs, IPMTs and other talent, high-quality meetings, strategy departments and reserve teams are the muscle.
After combining the views of consulting firms, current Huawei employees, other companies’ practices and his own pain, 刘靖康 believes the skeleton should be reduced to match the company’s capabilities at its current stage. Rash organizational restructuring can “hit a major artery,” while the company must also manage former cadres’ expectations, correct misaligned hires and clarify authority.
26. Systems Protect the Floor; Culture Handles the Gaps and Raises the Ceiling
刘靖康 says culture is not measured by slogans on the wall, but by which principles become a shared understanding when interests conflict. Practice by the person at the top is necessary; what tests the culture is whether the organization acts decisively when someone violates its values.
He believes Insta360 currently has no obvious fiefdoms or office politics, making organizational adjustment relatively simple. But whether a miscast person in a new role can be corrected quickly still tests whether managers are willing to take real responsibility.
“Responsibility without boundaries” addresses corner cases that systems cannot cover. Even with clear responsibilities, gaps will remain; if everyone guards only their own boundary, no one ultimately takes ownership. Identifying a system problem and escalating it proactively is a higher level of responsibility.
Culture cannot make a bad system work through “love.” Insta360’s sequence is to improve processes, incentives and the talent pipeline first, then let values filter down from senior executives, first-level department heads and managers, because a direct leader’s behavior is more real than a slogan.
27. The Challenge in AI Editing Is Not the Demo, but Delivering the Desired Result at Low Cost to the User
Insta360’s automatic-editing productization was around 60 points a few years ago and, by its own assessment in 2024, still below 70, with a target above 80. 刘靖康 emphasizes that automatic image and video editing depend on product design as deeply as on the model.
Customers do not really want “automation” itself. They want the content they need with as little effort as possible. The system must infer user intent and choose among on-device accuracy, performance and cloud cost.
Agents are making the team rethink interaction. Traditional UI becomes capable of fewer functions as it gets simpler, while more functions can intimidate novice users. In theory, natural language asks only that users be able to state their needs, which could materially lower the editing barrier.
Insta360 has invested continuously in automatic shooting and automatic editing since around 2018 and 2019, with 刘靖康 spending roughly 10%-15% of his time on combining AI and products. He does not view CapCut as a direct rival: CapCut serves a broader user base and improves the Insta360 customer experience, while Insta360 differentiates through proprietary hardware, cloud services, automatic shooting and scene understanding.
28. The Common Trait of Century-Old Companies Is Focused Direction and Continuous Capability Inheritance
During supplier visits to Japan and Taiwan in 2019, 刘靖康 saw history walls spanning 40 or 80 years, while Insta360 was only 4 or 5 years old. The depth of time led him and his colleagues independently to set the goal of building a century-old enterprise.
He observes that long-lived companies may change their main business every 7 or 8 years, or every half economic cycle, but capabilities are inherited and deepened across businesses. The direction remains relatively focused, with the second curve gradually becoming the first.
Insta360 is currently more interested in studying Sony. 刘靖康 calls it “dutiful” and “adorably straightforward”: it develops technologies and products the market may not buy. Like Nintendo creating new ways to play, it accepts the risk of failure to create a market rather than merely competing for time within an existing one.
Insta360’s product roadmap already contains 3 fully original, world-class targets. 卫诗婕 points out that roughly 63% of Sony’s profit comes from its insurance business, providing cash to fund its willingness to experiment. 刘靖康 acknowledges that Insta360 has no comparable cash cow, but says “using the devil to train yourself is not a bad path either.”
29. 刘靖康 Wants Investors to Buy the Stock Like Writing a Love Letter, Not Chasing a Concept
On speculation in “Insta360 supplier concept stocks,” his response is: “Please, please don’t get harvested by the concept.” The P/E ratio represents expectations; when short-term expectations are pushed too high, company growth often cannot catch up, and a shift in consensus can hurt buyers at the top.
Borrowing a phrase from Hanzawa Naoki, he says every company’s stock is like a love letter written to the company. The ideal investor likes the product, understands the opportunity and sees the flaws and challenges, rather than chasing something merely because it looks attractive on the surface.
Insta360’s track record makes the market prone to optimism, but new businesses require more time to validate. He wants investors to “be more optimistic about the long term and more objective about the short term,” using customer satisfaction as a simple indicator.
If large numbers of customers are dissatisfied with the product and service, investors should be cautious. If customers remain satisfied, confidence can rise. 刘靖康 has no stock account and no investment record, and was even unable to participate in the strategic placement because of it; he does not hide his unfamiliarity with the secondary market.
30. As a Public Company, Insta360 Will Still Take Risks—but Liquidation Is No Longer an Acceptable Outcome
The listing increased cash reserves, follow-on financing capacity and the available tools for employee incentives, while also placing more investor trust on the company. Management knows far more than the market, so major decisions cannot always be backed by external consensus.
刘靖康 admits that before listing, he could consider going all in on profits and saying, “If we bet wrong and the company gets liquidated, so be it.” Today, he cannot make the same bet with investors’ money; the floor for decision-making must rise.
That does not mean abandoning large investments. It means using more granular plans, more conservative preparation and hedges to limit the worst outcome. A company that makes no risky decisions cannot capture the high returns that technology may offer.
He is especially concerned about a public company’s “not knowing what it doesn’t know.” In the past, the company alone paid for mistakes; now secondary-market investors can be affected as well. Legal compliance and information gathering for major decisions must therefore be more cautious than before.
31. The RMB1.938B Raised Will Move Manufacturing from Delivery Execution into R&D Capability
Insta360’s R&D expense ratio has remained above 13% for years. 刘靖康 sees the company and investors as making a two-way choice: technology companies naturally offer high risk and high return, and investors unable to accept a single failed investment may simply be unsuited to this kind of long-term path.
The RMB1.938B raised will mainly fund an intelligent imaging device production base and a Shenzhen R&D center. A thicker post-listing cash position will also increase the company’s tolerance for R&D investment risk.
Manufacturing construction covers both pilot production and mass production. Pilot production requires iteration on processes, production-test equipment, fixtures, line design, first-pass yield and overall yield. Together, these variables determine product cost and time to market.
Insta360 started with software, then moved into hardware and deeper manufacturing. The Zhuhai base is intended to shorten yield ramp-up and line-optimization time for complex products, turning supply-chain capability into part of product development.
32. Insta360 Rode Four Tailwinds over the Past Decade: Capital, Content Platforms, Supply Chain and Logistics
When 刘靖康 had just graduated, the “mass entrepreneurship and mass innovation” era gave young people a chance to obtain venture capital. Without key financing rounds to keep it alive, Insta360 might not have crossed the most fragile phase of its early years.
In 2016, YouTube entered a golden period for content. Influencers and opinion leaders expanded rapidly, creating low-cost global marketing for vertical hardware. Traditional store-by-store placement, television and offline advertising were poorly suited to a startup’s need for precision.
The maturing smartphone industry brought dividends in components, processes, supply chain and talent, allowing each product generation to improve sharply while prices moved only modestly. During the pandemic, cross-border logistics and e-commerce adoption further lowered the barrier to overseas delivery.
刘靖康 acknowledges that if any one link had been missing, the company might not have made it to today—or would have needed far more time. Going global from day one was partly a choice and partly the infrastructure provided by the era.
33. The Ace Pro 2 Fogging Recall Rewrote Quality Management from Outcome Metrics to Preventive Process
刘靖康 calls the fogging issue at the Ace Pro 2’s initial launch the most sweat-inducing moment of the past decade. After tens of millions of RMB in R&D investment, the product was halted and recalled immediately after launch, and he briefly feared that the entire generation or even the product line could end.
The team first cleared the backlog: apologizing to customers, replacing units and issuing refunds, then reorganizing the team, quantifying the gap with competitors and solving the product and market problems step by step.
In March this year, thorough research showed that the product also worked for street photography, not just sports. The team developed a handle and cage, turning it into a miniature mirrorless camera with attractive “Leica color.” During “618,” the new product repeatedly sold out, prompting consumers to complain about hunger marketing.
刘靖康 accepts responsibility: both product R&D and quality were under his name, but he had managed only final quality outcomes rather than systematically managing talent reserves, SOP implementation, supplier materials and preventive mechanisms. The system still has gaps, but closing them is now a firm priority.
34. The True Organizational Inflection Point Was the Team Getting the Product Right Without the Founder’s Approval
The Ace street-photography kit was developed with no involvement from 刘靖康; he did not even know about it until launch. Had he reviewed it, he might have rejected it for being too heavy and inconsistent with his personal aesthetic. After it succeeded, his lesson was: “This company may be able to make money even without me.”
Another source of excitement was the team’s decomposition of its short-video-platform gap into more than 20 accounts and content metrics. Lagging metrics were marked red, large gaps dark red and leads green. The board gradually shifted from all red to more than half green, like “reclaiming lost territory.”
The team’s management principle is “celebrate small wins.” Any improvement in a quantified metric earns timely recognition; the team does not wait for the final battle to end. 刘靖康 believes participants need continuous positive feedback that problems are being solved and advantages are being built.
These changes confirmed that the context not control direction was working. Talent is far more complex than imagined, but the payoff is also far greater. He saw that the founder need not be a necessary condition for every success—the evidence that delegation is truly working.
35. Anonymous Truth-Telling and a Diamond Analogy Form Insta360’s New Talent Contract
At one team-building event, employees anonymously voted red or black cards. Around 30% had considered leaving in the prior 6 months; roughly 60%-70% thought the boss “was an idiot”; around 60% believed the handheld product line could surpass the rival in a year. Anonymity let management see the truth, and let employees know that management had seen it.
For the final question, each employee who voted for confidence received an immediate RMB1,000 raise; if the target was achieved a year later, they would receive another RMB3,000, while those who had not believed received RMB1,500 as well. The design rewarded confidence immediately while using a long-term target to unite people with different views.
刘靖康 often says a 1-carat diamond costs about $10,000, while a 10-carat diamond is not $100,000 but could be $1M: “Talent is priced by scarcity.” Scarcity comes either from taking one thing extremely deep or from becoming a cross-disciplinary generalist.
The company has begun systematic talent mapping, cross-level one-on-ones and rotations across more than 3,000 people and over 600 job types. The shared goal is not to guarantee that employees never leave, but to ensure that after 3 or 5 years at Insta360, they command a higher market price even if they do.
36. The Most Realistic Way to Retain Top Talent Is to Let Them Leave in an Orderly Way
刘靖康 believes top talent will probably start companies of their own. DJI could not retain 陶冶, who later founded Peking. Rather than assume people can be kept forever, he would design the company as a platform where they acquire knowledge, experience and resources.
The heads of Insta360’s virtual product lines resemble internal first-seat executives. They can practice team building, product and operating decisions, while the company as a whole absorbs the cost of mistakes; other product lines can also share lessons from incidents such as the Ace Pro 2 fogging problem.
If a leader builds a business and team that perform just as well—or better—without him, and the project wins recognition from outside investors, Insta360 can co-invest in his startup. The company may even allow him to take some colleagues, provided he first finds a better successor.
The openness has boundaries: a new business cannot compete with Insta360, and the talent must communicate honestly. 刘靖康’s logic is that prohibitions cannot eliminate colleagues starting companies together; negotiating in advance and arranging succession can reduce the real loss.
37. “Intelligent Imaging” Is Not Adding AI to Traditional Cameras, but Finding Scenarios Users Have Not Yet Articulated
刘靖康 defines traditional imaging as meeting existing needs more efficiently, relying mainly on mechanics, optics and semiconductors. Intelligent imaging depends more comprehensively on software, AI and algorithms, and places greater emphasis on creating new scenarios.
Imaging demand is an open world. Consumers often “do not know what they want until you give them the solution.” Insta360 therefore does not only optimize known categories; it also tries to discover new ways to capture and share that have not previously been expressed.
Compared with passively following specifications and marketing moves a year ago, 刘靖康 now emphasizes proactive design. Starting from why customers choose Insta360, the company asks how to make its PMF judgment more accurate and efficient than competitors’ over time, ultimately bringing the question back to organization, process and talent.
The encouragement from DeepSeek and ByteDance is twofold: Chinese companies can reach world-class quality, and small companies may win through first-principles thinking and steady incremental progress. But he also warns that domestic rivals are often more comprehensive than international ones, making the competition potentially much harder.
38. The New York Queue Proved Five Generations of Product Accumulation—and Exposed the Need for Deeper Offline Brand Building Overseas
On April 22, Insta360’s new 360-degree product went on sale in New York, drawing a longer-than-expected line outside Grand Central Terminal. It was the company’s first offline launch of this type; previously, it had relied more on direct listings through Amazon and its own website.
刘靖康 says the lesson was not simply that the product was popular, but that 5 consecutive generations of 360-degree products had built a loyal user base. A similar event in Tokyo might have produced a comparable result.
Some overseas consumers already see Insta360 as a brand representing China, but company research shows awareness still trails competitors clearly in many markets. Marketing touchpoints have historically been mostly online; the next stage requires entering the deep water of stores and offline experiences.
39. An Overseas Grain Bin and a Chinese Test Kitchen Form a Two-Way Growth Flywheel
On concerns about its high overseas revenue share, 刘靖康’s direct response is: “If you’re worried, don’t buy our stock.” He will not raise China’s share simply to make the structure look better; business allocation remains driven by opportunity and efficiency.
China understands target customers better, and its marketing infrastructure is more advanced. He believes Douyin is more mature than TikTok and Meta, making China an ideal place to validate marketing practices before replicating what works overseas.
The full sequence is: diversified overseas revenue creates a grain bin that funds high-intensity experimentation and provides a margin for error in China; higher efficiency developed in China then feeds back into overseas markets, expanding the grain bin and the room for the next round of experimentation.
Domestic competition is broader and execution more extreme, so the highest intensity in the near term will likely be in China. Long term, however, the company must return to the global stage and learn from leading companies’ board oversight, power structures and CEO succession mechanisms.
40. Competition Will Ultimately Move from Products to Ecosystems, Organizations and Talent Supply Chains
刘靖康 describes the evolution of competition in layers: first technology and products, then marketing, supply chain and service, followed by efficiency and intelligence. At the higher levels come ecosystem, organizational design and the talent supply chain.
Ecosystem competition includes forcing channels or suppliers into a binary choice. Core suppliers are generally large and hard to coerce; non-core partners may side with a giant if it contributes 20%-30% of their revenue. Insta360 understands that commercial self-interest.
The solution is not merely to fight over existing partners, but to grow with suppliers like Huawei and Apple do, investing in talent, capital and process capabilities. 刘靖康 cites Hongjing Optoelectronics’s listing as an example, saying both sides developed through cooperation on lenses and modules.
The same applies to talent. Technology companies need to develop new knowledge and solve problems no one has solved before, so the market often has no ready-made candidates. Competition moves upstream to cultivating and attracting students in their first or second year of university. Even when a graduate Insta360 wants accepts its offer, Huawei or another large company may still rank higher on the person’s priority list.
41. Compete Across Every Dimension Tactically, but Never Become the Rival’s Shadow Strategically
Insta360 says its revenue in the first half of 2024 was already slightly above GoPro’s, while 刘靖康 does not deny that DJI is growing rapidly. Every link—product, marketing, supply chain and organization—may be necessary, and any weakness can drag down the final commercial result.
Over the next 3 to 5 years, Insta360 wants to become a more complete “12-sided fighter,” while creating new industry practices in marketing, supply chain, talent incentives and culture. The highest praise he wants is: “So this can be done this way too?”
But if strategy focuses only on the rival, the company can become nothing more than its shadow, another version of it or a smaller version of it. Only by moving the question up one level can Insta360 gain an advantage while finding an exit from homogeneous competition.
42. Imaging Is an Open World, So Creativity Can Offset a Giant’s Funding Depth
Asked about the label “small giant,” 刘靖康 first corrected it: “We’re not a small giant. We’re a big head—the head is just very big.” He then reframed the question as how a “big head” competes with a giant.
He is glad the competition is taking place in imaging rather than an industry where standards fully converge. Autonomous driving may ultimately converge around safe arrival, speed, minimal intervention and comfort, leaving efficiency and price as the remaining battlegrounds.
Imaging is more like a game: every product can establish its own world and create incremental demand through unprecedented experiences. Efficiency still matters, but creation comes first, providing products with pricing power, gross margin and funds for the next round of R&D.
That is why he calls “killing on price an act of laziness.” If Insta360 can keep growing through creation despite having a shallower pocket than the giant, it would prove that markets need not all degenerate into resource-saturated attacks. Otherwise, innovation may disappear.
43. “Employees First, Customers Second, Shareholders Third” Is Insta360’s Core Experiment in Competing with DJI
Asked about the most fundamental difference with DJI, 刘靖康 did not cite youth or energy. He offered an organizational order instead: “Employees first, customers second, shareholder returns third.”
The classic shareholder-first model can sacrifice customers for short-term returns. Insta360’s proposed virtuous cycle is to first develop people who can create knowledge, then have them deliver incremental market value to customers, ultimately generating shareholder returns above those available from competing for existing demand.
Profit from incremental value can then fund training, experimentation and research, creating greater tolerance for error. He uses the rival’s employee turnover as an indirect reference and believes it is difficult to prove the rival puts employees first, while acknowledging that Insta360 is still only a “big head” and cannot prove in advance that the experiment will succeed.
Youth does not directly create an advantage; experience shows up faster in competitive results. The long-term value of youth lies in curiosity, openness and the possibility of change. The difference between 30 and 40 may be small, while the difference between 40 and 50 may gradually become visible.
44. Before Employees Love the Company, It Must Solve Why They Would Stay
刘靖康 breaks the foundation of retention into 3 categories: good material compensation; encouragement and a sense of achievement through challenge and failure; and continued growth, rotations and opportunities to build composite capabilities.
Employees do not need all 3. Satisfying 1 may be enough to keep someone. If Insta360 can give large numbers of employees 2 or even all 3, he believes there would be “no reason not to love this company.”
Paying more is the biggest common denominator, but sustaining it requires high operating performance, and it is not even necessary for some senior talent. Getting employees broadly to love the company remains a proposition that must be “full of challenge and full of inspiration.”
Think Bold’s official Chinese translation is “始于敢想,” but 刘靖康 plans to distill the values again. He stresses that uniqueness cannot override science: the company can adopt unconventional practices, but they must still follow business logic and remain coherent over the long term.
45. Interests with No Obvious Payoff Eventually Connected into Insta360’s Startup Path
In high school, 刘靖康’s focus on programming hurt his grades in the first 2 years, before he returned to normal study in the third year and entered Nanjing University. He also mentions the 20 points from a competition. The more important positive feedback was that taking a different path from most people looked like a detour, but the abilities he truly loved eventually mattered.
A failed pitch in his sophomore year taught him that impressive technology does not automatically create a commercial scenario: “Business is a very serious matter.” The imaging startup was not planned on day one; it grew out of the live feeling of campus music festivals and a desire to find a better way to record and share it.
The team initially worked on live streaming. After seeing 360-degree video, it confirmed that the right product was a 360-degree camera. Software-engineering experience, campus scenarios and the impulse to record gradually connected into a single path.
One month before graduation, IDG’s 李丰 decided to invest after roughly 40 minutes. 刘靖康 believes the investor was mainly betting on his university experience and growth trajectory rather than the project itself. If the decision had come 1 month later, the company’s cash chain might already have broken by graduation.
46. “You Only Live Once” Became the Hedge That Keeps Him Brave After He Has More to Lose
刘靖康 once said he was brave because he had nothing. Now he acknowledges the basic human fear of loss. With obligations to employees, customers and investors, he obviously cannot take risks as he did when he had nothing.
His solution is not to deny the debuff, but to find something he fears losing more: life happens only once. If the goal is merely to become larger than the rival or hold more cash, the result will be another similar version.
“Anyway, you can’t take anything with you when you leave.” He cares more about doing what no one else has done, making the journey sufficiently vivid and leaving without regret: “you only live once.”
The listing raises constraints but need not produce pessimism. Public-company responsibility protects the floor, while a finite life continues to push the ceiling. Together, the 2 create a more complicated form of courage than having “nothing to lose.”
47. He Still Wants to Build Products, but Has Had to Learn to Become an Entrepreneur
When asked in 2022 to choose between being a “product creator” and a “broad-minded entrepreneur,” 刘靖康 chose the former. His answer had not changed at the time of the listing, but in reality much more of his time had shifted toward building the company.
He compares management to preparing for an exam. Playing games delivers happiness faster, while getting into a good school requires night study and practice tests. He may not enjoy the process of fixing the organization, but he badly wants the result and does not assume he will necessarily come to love management later.
Asked whether the company had had it too easy, he gives 2 answers. By induction, almost every company eventually faces major adversity. From first principles, if every new problem receives the right decision, there is no reason negative feedback must follow.
The real risk is that past success reinforces path dependence, leading the company to use old experience to answer future problems. The countermeasure is self-reflection and iteration across countless micro-decisions, relying less on the past and continually updating the method of thinking itself.
48. The CEO Decision Rule Shifted from “One Objection Stops It” to “One Strong Conviction Makes It Happen”
Facing more intense competition, 刘靖康 and the CEO merged their offices to increase information flow and the frequency of joint decisions, while reallocating work that had fallen into gaps in authority and responsibility.
In the past, when they disagreed, one person’s strong objection was enough for the other to stop pushing. Now, if either person strongly insists, the other does not object, provided they have first exchanged information fully.
The reason is that as the business expanded, the 2 executives saw very different sets of information. The person insisting probably holds more facts and logic, and is willing to take responsibility for the decision. “Whoever insists decides; whoever decides takes the blame.” The person objecting does not have to take the blame for a missed opportunity.
His reminder to himself is “don’t regret too easily.” Major decisions were carefully considered at the time and should not be overturned because of short-term volatility. The next ambitious targets remain world-class original products and an organization built around employees first, customers second and shareholders third.
49. What Insta360 Must Not Lose After the Listing Is Its Fun, Rebelliousness and Independent Judgment
刘靖康 says the thing he finds hardest to accept is: “You only live once, yet you go and follow other people’s ideas.” There is no problem if independent thought happens to align with consensus. The real betrayal of oneself is reaching a different conclusion and submitting anyway.
He does not promise to remain forever in a youthful state, but hopes the CEO identity will not completely cover up 刘靖康 the person. When he sees a rainbow, he still wants colleagues to stop working; he is still willing to bet on high-risk designs in gift boxes, and still willing to express fear and negative emotions publicly.
His Think Bold for the next decade contains 3 unfinished propositions: make world-class original products, build an employee-first company and use a rival like DJI to train Insta360 into a more complete organization.
“Building a company that strives is already hard; building a company where everyone strives is even harder.” The half-joking close preserves his contradiction: he does not enjoy every entrepreneurial responsibility, yet still wants to build a company in a way no one else has.