124: Ex-Li Auto President 沈亚楠’s $500M Bet on Middle-Class Housing

Summary

• 沈亚楠 has invested roughly $500M in 栖息地, betting that China will eventually produce a “Chinese Home Depot” with RMB1T in annual revenue and a 20% share of the RMB5T home-improvement market. He acknowledges that the company is about three quarters behind schedule and has cut its annual bookings target from the 10,000s to the 1,000s, but still believes a short supply chain and a complete product will ultimately win: “I hope I’m the person who finishes the 42-kilometer race, not the most tragic person who collapses at kilometer 40.”
• 栖息地’s core bet is not to renovate homes better, but to replace “hand-assembled renovation services” with an integrated smart home through its RMB399,800 L32. 沈亚楠 describes the model as “modular R&D and manufacturing, customized combination delivery”: centrally controlling air conditioning, fresh air, water purification, lighting, furniture and the smart hub, while sourcing only a small number of core appliances externally. The aim is to reduce coordination across trades, SKU sprawl and supply-chain waste, so customers know exactly which modules and functions they are buying when they pay.
• The February 28 launch showed that a good product does not automatically create demand: online exposure may have set an industry record, but 栖息地 expected tens of thousands or more visits and inquiries and got only thousands—an order of magnitude fewer. An A/B test that withheld the price did not perform much better, suggesting that RMB399,800 is not the primary obstacle. The real problem is that users are not yet interested enough to hear the full logic of air, light, water and intelligence: “If he isn’t interested in you, why would he spend time understanding your completeness and complexity?”
• 沈亚楠 has therefore revised the product-first thinking he brought from Li Auto: complex purchases first take shape through “System 1,” then use “System 2” to find the reasons. 栖息地 used to lead with health, comfort and technical specifications; it now plans to put “beautiful relationships, natural elegance, health and comfort” at the front of the brand in that order. Technology and phrases such as “L3-level household AI Agent” will remain, but they are “not enough, and may not even be the most important part.”
• The Chinese Home Depot in 沈亚楠’s vision is not 2,900 building-material superstores, but a platform combining proprietary products, central-warehouse logistics, online channels and decentralized delivery partners. Home Depot operates in an approximately $800B U.S. Home Improvement market with about a 20% share; 栖息地 wants to provide either a complete smart home or a “chassis” built around air, light, water and intelligence, then let local contractors, studios and construction crews handle the last mile. “Having one person acquire the customer, then having that chain leader coordinate the R&D process for complex products, is the most efficient way.”
• A downturn in home renovation first tests cash discipline and delivery capability, so 栖息地 places customer deposits in segregated accounts, with unused funds refundable even if the company stops operating, and sets aside separate after-sales reserves. 沈亚楠 considers this a basic corporate obligation. On delivery, the company is moving from employing its own workers to a model combining training, certification, quality control and crew incentives; the key is not just craftsmanship, but ensuring the right, complete materials arrive at the right time.
• Whether the company can scale will determine whether this venture becomes a product company or an expensive long-term experiment. 沈亚楠 estimates that roughly RMB2B in revenue would allow operating cash flow to support the R&D needed for his vision of continuous innovation. He wants to see the smart-home product he imagines no later than 2028, including cabinet-appliance integration, household-item recognition and service robots, with 栖息地 solving the “scenarios and brain” while embodied-intelligence companies solve the “actuators and cerebellum.” The biggest risk today is not that the direction is unreasonable, but that “it is not accepted within your timeline.”

Deep dive

1. 沈亚楠 is betting the next 10 to 15 years on being a founder at the top

• In summer 2022, 沈亚楠 was recovering at home after an injury when he finally stepped away from the relentless workload at Li Auto. He asked himself whether he could work another 10 to 15 years. “Isn’t life just an experience?” In the end, he chose to test his limits again: he had already co-founded a company from the number-two position, so the next step was to take on more of the risks and pressure as the number-one founder.
• 栖息地 was almost the only direction that emerged. 沈亚楠 had just renovated a home, was already an architecture and design enthusiast, and even described himself as more interested in housing than cars. The market was large and durable, with room for technological evolution—meeting his threshold for starting a company.
• When 程曼祺 asked why he would not take a break, he replied, “Rest isn’t exciting.” A worsening macro environment, in his view, was a reason to start earlier. For major decisions like this, he “rarely consults anyone”; visits to veterans of the home-renovation industry were for learning how, not for holding another vote on whether to do it.

2. Three years of R&D produced the first product—and a market-education result below expectations

• For its first six months, 栖息地 was led by “老姜,” who assembled the team and developed the product and business model. The company then spent a full 2 years developing L32 and validating it with angel users. On its third anniversary, it officially launched its first “flagship three-bedroom smart home” on February 28, with a standard version priced at RMB399,800.
• 沈亚楠 describes entering a “24-hour workday”: nearly all his money, time and energy have gone into the company, and “this is extremely difficult, yet the result is still below expectations.” The veterans who once warned him away from home renovation were at least “telling no lies” about the difficulty.
• He compares the current stage to the first 4 months after the launch of Li Auto’s Li ONE, not to a renovation company that opens its doors and immediately starts generating profitable orders. The supply-chain, product and R&D costs of the first 3 years have already been incurred; sales must ramp quickly, or it will be like “building a car and not selling it.”

3. L32 is trying to eliminate “hand-assembled renovation,” not a particular design style

• 沈亚楠 once worried that users would not understand the company, so he compromised by saying 栖息地 also did “renovation.” He now believes that compromise created conceptual confusion. “Language is the boundary of thought.” No one asks whether Xiaomi can install a Gree air conditioner; 栖息地 is not selling a service that combines arbitrary branded components.
• The new definition is “integrated smart homes”: customers buy a complete product made of multiple compatible modules, while 栖息地 handles R&D and coordination across the product. Its method is “modular R&D and manufacturing, customized combination delivery,” not redesigning every home from scratch by hand.
• Traditional renovation sells customers an aspiration and desire for a beautiful home when they pay. Contractors, meanwhile, face different ceilings, toilets, cabinets and brands every day, so the result depends on inherited craftsmanship rather than repeatable products. 沈亚楠’s question is: “It isn’t a product, so there is no certainty. How can you be satisfied?”
• 程曼祺 correctly points out that “integrated smart home” is still difficult for an ordinary user to visualize quickly. 沈亚楠 acknowledges that users do not naturally care whether they are choosing an integrated smart home or an assembled service; before entering that comparison, they first want to know, “Who can I trust?”

4. Home Depot’s real value lies not in the superstore, but in products and supply-chain capability

• 沈亚楠 uses Home Depot to explain 栖息地’s end state: U.S. Home Improvement—“home improvement” in the broad sense—is an identifiable market of roughly $800B, and Home Depot holds about 20%. The program notes that the $400B market-cap figure he cited is recently closer to $365B and reached roughly $440B in the second half of 2024.
• In his view, Home Depot appears to be a building-material superstore but is fundamentally a product company first: roughly half its products are developed, sourced and even designed in-house. Even external brands are selected and enabled by the company, making the logic closer to Private Brand retail than simply taking a channel margin.
• Its second layer is logistics, warehousing and ecosystem support. Home Depot has roughly 2,900 large stores across the U.S., aiming to put materials within 10 miles of 95% of the population while also offering tool rentals and construction training. The stores are forward warehouses first, and display and transaction venues second.

5. China’s Home Depot will not copy U.S. stores, but it will copy the platform logic

• The U.S. Home Depot was born in the “assembly era” of 1977, when selling components was enough. 沈亚楠 believes China is moving into the era of integrated smart homes, so 栖息地 could sell complete homes or provide ecosystem partners with a residential chassis integrating air, light, water and intelligence.
• China does not need 2,900 big-box stores to cover pickup demand. 沈亚楠 envisions central warehouses in Beijing, Shanghai, Guangzhou and Shenzhen, with delivery and online commerce providing coverage. The platform would look more like a Private Brand e-commerce and logistics system serving proprietary products.
• Delivery also need not copy the U.S. DIY model. In its early days, Home Depot would draw plans for a customer’s chicken coop and list the wire mesh, stakes and lighting, turning middle-class people who could not design but could work with their hands into DIY users. China’s existing logistics and professional-construction networks make certified service providers the more likely last-mile solution.

6. Mixue Bingcheng and the auto industry show that short supply chains work beyond housing

• When explaining “small orders, fast response” to the internal supply-chain team, 沈亚楠 has them study Mixue Bingcheng: headquarters controls ingredients, the supply chain, the brand and the complete beverage, while the couple-run store handles the last mile. “Isn’t a cup of lemonade a complete product? Isn’t it selling sugar and lemons?”
• The auto industry has the same structure: upstream component suppliers, automakers and 4S stores. The chain leader provides a complete product, acquires customers centrally and coordinates complex R&D. By reducing intermediate links, 沈亚楠 believes this “is the most efficient way, so it will inevitably win.”
• The difficulty has not weakened his confidence; it makes him feel that “the more I do it, the more flavor it has.” His concern is not that the direction is wrong, but whether he can finish the 42 kilometers: “I don’t want to be the most tragic person who collapses at kilometer 40.”

7. Industry shakeout first exposes insufficient value and diverted customer funds

• 程曼祺 notes that more than 60 renovation companies had suffered funding-chain failures in the first half of 2024, followed recently by the 住范儿 incident. 沈亚楠 declines to comment on individual cases and offers only this judgment: “When the economy is good, you can muddle through. When it is bad, why should you survive?”
• From the start, 栖息地 has placed customer deposits in segregated accounts and does not use them for payroll or daily operations. In the worst case, if the company stops operating, unused money for projects still under construction should be returned; delivered customers have separate after-sales funding arrangements.
• 沈亚楠 says this is not a selling point that requires special promotion, but a baseline any company should meet.

8. The launch secured industry-level exposure but generated one order of magnitude fewer actions

• The February 28 launch may have generated the largest online exposure in the industry’s history. 栖息地 expected tens of thousands or more visits and inquiries, but got only thousands—one order of magnitude fewer. Professionals, property developers and renovation companies understood the product; it did not break through to To C users.
• Complex products create a self-locking paradox: completeness should be a reason to buy, but users must first be willing to spend time hearing about air, light, water and intelligence. If they are not interested at first sight, they will not learn about completeness; without understanding completeness, they do not become interested.
• 程曼祺 infers that ordinary people may need “one feature, not completeness.” 沈亚楠 immediately agrees: after hearing “integrated smart home,” the most common response is still, “What does this have to do with me?”
• Price is not the main explanation. The team ran an A/B test without showing the price, and the result “wasn’t much better.” Whatever the quote, someone in the market will offer a lower number; the issue remains whether 栖息地 has first opened a cognitive entry point.

9. Product thinking must pass through System 1 before it earns a chance to explain the product

• 沈亚楠 believes Li Auto sells well not merely because of its styling, but because its brand strategy creates an overall feeling. The program then discusses Xiaomi SU7 and the dual openings in the front compartment of the SU7 Ultra. Once the material baseline is high enough, emotional value accounts for a much larger share. The “style over substance” judgment comes from 程曼祺’s example, not from 沈亚楠’s position on 栖息地.
• 沈亚楠 agrees that complex purchases are decided first by “System 1,” with “System 2” searching for supporting reasons. 栖息地 produced extensive System 2 material on air, water, lighting and intelligence, but lacked the emotional impact that opens the door at the beginning. System 1 is not blind impulse; it is a high-dimensional intuition formed by every brand touchpoint.
• 栖息地’s showroom already has a CMF Lab offering more than 200 materials and even more color combinations, but the launch barely mentioned it. Terms such as “CMF,” “smart chassis” and “the world’s first L3-level household AI Agent” strengthened the technology signal while raising the cost of comprehension. 沈亚楠 admits the team is “especially male-brained”: technology is necessary, but far from sufficient.

10. A “happy home” should improve relationships before adding equipment

• After reordering the value proposition, 沈亚楠 puts “beautiful relationships” first: when people recall their happiest home, they think first of the family, not the house itself. The role of a residential product is to use space and equipment to create more opportunities for people to be together.
• The open kitchen is the clearest example. A closed kitchen makes the person cooking face away from the family like a servant; turning the sink or stove toward the living room allows the cook to keep talking. For 沈亚楠, cooking while facing one’s family is what makes a family feel like a family.
• In the children’s study pod, the desk is supported entirely from the wall, with no table legs where mother and child sit side by side, so the person accompanying the child is not blocked. This small detail shows that relationships are not an abstract slogan; they can be built into furniture dimensions and structure.
• A low-noise stove hood reduces the sound of high-heat Chinese cooking, allowing communication between the kitchen and the rest of the home. The test for product innovation is not simply whether a parameter is higher, but whether it “serves this beautiful relationship.”

11. “Natural elegance” is both an aesthetic position and the result of storage and less housework

• The second layer of value is summarized as “natural elegance”: use more natural materials or materials with a natural tactile quality, pursuing ease rather than display. 栖息地 will not cover walls in luxury stone or marble, nor create “screaming design” to capture short-term attention.
• “Elegance” first requires refined execution, then a stable and distinctive aesthetic position. 沈亚楠 wants a home to remain current 10 or 20 years later, so he is cautious about trends such as microcement: it may crack, be difficult to maintain, and its popularity may not last like wood flooring that has been used for hundreds of years.
• Most households cannot yet achieve full decluttering, so elegance must be backed by sufficient storage. 沈亚楠 mentions an “800-cabinet” module for strollers, sports equipment and suitcases. Once belongings are scattered across the floor, even the most expensive visual design collapses.
• The “less housework” example is an integrated stove, cabinetry and countertop: move grease-prone, hard-to-clean knobs to the side or underneath, reduce countertop protrusions and boundaries, and free up food-preparation space. Only by controlling the stove, cabinets and countertop together can the cleaning problem be designed out at the R&D stage.

12. Health and comfort remain the product chassis, but no longer lead the communication

• Air, water and light form the third layer, “health and comfort,” and also the foundation: “If the air, water and light are not healthy or comfortable, what else is there to discuss?” Intelligence is not the endpoint; it serves health and comfort.
• 栖息地 previously reversed this order, leading with air environments, smart environments and dense technical specifications. Its provisional brand priority is now beautiful relationships, natural elegance, and health and comfort—establishing the feeling first, then explaining the engineering system behind it.
• 沈亚楠 stresses that this is not abandoning product thinking. The 3 values will not change; what changes is the order of communication and the design of touchpoints. The company needs to keep testing until it finds a way to communicate that users can perceive, trust and be convinced by.

13. Old floor plans are not a counterexample to standardization; modules can serve new living relationships

• 程曼祺 poses a challenge with an old Beijing apartment: the living room is tiny, the bedrooms are large, and load-bearing walls cannot be changed. How can an open kitchen and island work? 沈亚楠 cites a shipping-container customer who combined the former corridor-like living room and kitchen into a kitchen, then converted the largest south-facing bedroom into the living room without moving walls.
• The deeper question is whether one should “cut the foot to fit the shoe” or buy a shoe for the foot. If a family has only one child, it may not need to use all 3 rooms marked on the floor plan as bedrooms. Sleeping space can be smaller, while the largest, best-lit room should instead support interaction among the family.
• Wide variation in floor plans does not mean every component must be reinvented. 栖息地 breaks the product into Lego-like modules and customizes the combination on site. What changes is the arrangement of modules and the use of rooms—not a return to infinite SKUs and hand-assembled renovation.

14. A complete product must set boundaries; the ecosystem can absorb nonstandard aesthetics

• If a customer insists on a microcement 栖息地 has not developed or wants to buy only part of the construction service, 栖息地 can allow the customer to proceed independently, but can no longer guarantee overall environmental performance, warranty or schedule. 沈亚楠 accepts that this may cost the company the customer, because a complete-product promise cannot cover variables outside its control.
• Sensible new materials can still enter R&D, but the standard is not whether they are fashionable; it is durability, maintenance and system compatibility. Excessive individual variation would cause the supply chain, installation training and quality assurance to lose convergence again.
• Outside its directly operated complete product, 栖息地 can provide partner contractors only with a health-and-comfort chassis built around air, light, water and intelligence, while partners handle the exterior style. It cannot endorse every third-party choice, but it can control the foundational systems that most affect the living experience.

15. “Building homes like cars” refers to R&D and design, not turning construction sites into assembly lines

• A modern cabinet may connect simultaneously to walls, ceilings, access panels, refrigerators, electrical circuits and smart systems. It is no longer furniture that a single carpentry trade can handle independently. 沈亚楠 believes the developer of the complete product must coordinate the components, rather than leaving site designers to assemble them ad hoc.
• A recessed light channel shows how repeated manufacturing improves quality. Early workers could not build it; even after the first training, they were not fast enough. Once roughly 60% of users repeatedly chose the same ceiling, a contractor could build it both quickly and well by the third repetition. The team also lengthened each light segment from 7 meters, reducing uneven light caused by corners and joints.
• Except for a small number of appliances such as refrigerators, washing machines and robot vacuums, 栖息地 says it controls the design of most air-conditioning, lighting, water-environment, smart-system, fixed-furniture and movable-furniture products. Manufacturing remains outsourced to specialized factories, similar to Li Auto co-developing components with Valeo rather than building every factory itself.
• 沈亚楠 clarifies that “building” in “building homes like cars” mainly means R&D and design. Upstream suppliers manufacture components; 栖息地, as the middle layer, defines the complete integrated smart-home product; downstream owned or partner channels handle design and delivery.

16. The construction site’s hidden bottleneck is not craftsmanship, but people and materials failing to arrive together

• 沈亚楠 believes a worry-free delivery experience first depends on product convergence, with the second key being “people-material coordination.” Common delays do not happen because the contractor cannot do the work, but because the contractor arrives before the materials, or the toilet arrives missing one fitting, forcing every subsequent process to be rescheduled.
• 栖息地 currently serves roughly 400,000 users and has tried to select and train more mid-to-high-end contractors. The real difficulty is the logistics system behind them: getting the right things to arrive when they should, and making sure they are “right and complete.”
• A product company can build this system more easily because it has fewer SKUs, stable specifications and transparent supply-chain nodes. When renovating his own home, 沈亚楠 received daily calls from tile and other suppliers and personally coordinated the foreman and materials. Asking an ordinary family to play that role would consume essentially all its energy without guaranteeing the result.

17. Construction sites should manage crew outcomes, not pretend to manage every individual

• Traditional renovation companies “contract but do not manage.” Construction crews operate efficiently, but lack the ability to train workers and deploy new tools over the long term. The industrial-worker model turns contractors into employees, enabling standardized training but creating fixed labor costs, business volatility and on-site management expenses.
• Home renovation is also unlike an auto production line. A carpenter’s poor ceiling work may not become visible until painting, while tilers often work through flexible divisions in which one craftsman trains apprentices. It is difficult for a company to know exactly what each person did; the true tradable and assessable unit is the construction crew and its final output.
• 栖息地 is therefore keeping training, certification, quality control and recognition systems, but no longer insists that everyone be an employee. The incentive comes from “dividing the money correctly” and long-term recognition: good work increases income and creates a continuing incentive rather than a one-off game.
• Specialized roles such as smart-hub commissioning remain with in-house engineers; ordinary construction can be delivered by certified crews. The key is not whose ID card the worker carries, but whether the worker has genuinely learned the product before arriving on site.

18. The service cycle is 10 to 15 years, but early delivery errors immediately consume trust

• 沈亚楠 says negative construction posts on Xiaohongshu have fallen significantly and delivery strategy is gradually converging. But the platform still mainly shows renovation diaries and photos of empty homes after completion, making 栖息地 look like a renovation company and “like a high-end hotel,” without the lived-in warmth of a home.
• He wants more users to show long-term use, because a smart home serves 10 to 15 years of living, not 3 to 6 months of construction. Some users say 栖息地’s service resembles the service they expect after buying a car; such firsthand testimony is closer to the product’s value than construction-site traffic.
• What makes 沈亚楠 feel most guilty is when a certified contractor “could have done it well, but simply did not.” One early customer had problems with both fresh-air design and construction, so the entire ductwork system had to be removed and rebuilt at 栖息地’s expense. The customer was eventually satisfied, but the lost time before moving in could not be recovered.
• Another user expected a construction manager to serve only them, unaware that one manager might be handling 10 projects at once. The user had time to ask questions only at 11 p.m., and the manager might not respond. 沈亚楠’s conclusion is that “the industry itself is wrong”: explanations should be codified in documents, videos and proprietary data, with dedicated large language models trained where necessary, so the product serves the user rather than endlessly adding staff.

19. The delivery platform is opening up, but partners must pass the “3 rules of discipline”

• The launch did not break through to C-end users, but it attracted property developers and renovation companies. 栖息地 expects to enter at least 3 cities in the second half of the year through partner delivery rather than its own teams. Working with renovation companies is still essentially joint To C; property-development projects require a carefully defined To B boundary.
• Property-development business cannot tie up 栖息地’s cash, and the product brand must be visible. If 栖息地 merely hides behind the scenes as an anonymous supplier, the long-term value to the company is limited. It is not categorically opposed to full-renovation projects; it will decide based on brand, cash and product-control requirements.
• Partner contractors must meet 3 rules of discipline: they must be regional leaders that care about their reputation; they must build a dedicated professional team for 栖息地 rather than sell the product occasionally; and they must have a founder still working on the front line who can communicate candidly with 沈亚楠. Complex delivery is still “a project between the 2 people at the top.”
• 沈亚楠 believes a strong local construction company can be a good business if it understands the essence of delivery. 栖息地 does not need to replace such companies; it can amplify their capabilities with products, training and logistics.

20. Of the 3 strategies, the hardest challenge has shifted from product and delivery to brand

• 栖息地’s 3 strategies are product innovation, brand reputation and a service-provider platform. The product foundation is in place and new products will follow; delivery was poor early on, but the path is now relatively clear, making execution and talent the next challenges.
• “Building homes like cars” makes product innovation the team’s most familiar capability. Companies such as 爱空间 and 方林 have also shown that worker-management problems can be solved. The unresolved problem is how to make consumers understand a new product and a new category.
• 沈亚楠 admits the launch followed automotive playbook: create a sharp point and support it with online and offline activity. But cars only need to highlight a difference on an existing iceberg; 栖息地 first has to build the entire iceberg. Neither the format nor the content made users feel, “This has something to do with me.”
• He believes a rebuilt brand could plausibly generate monthly leads in the 10,000s; whether it can reach 100,000 is a question. Even failure means continuing to iterate, because the creator of a new category bears the highest education cost and usually has a chance to become the default brand.

21. Trust must be built by the founder, R&D capability and real homes

• 沈亚楠’s current view is that a user’s first step is not choosing a suit or sportswear, nor choosing an integrated smart home or hand-assembled renovation. It is finding someone trustworthy. With companies frequently disappearing, “I’m here, so everyone can rest easy” may be basic, but it is his direct advantage over most renovation companies.
• A higher level of trust requires evidence: the founder’s continued investment, R&D and factory capabilities, professional designers such as 卢老师, and homes that have actually been delivered. 程曼祺 offers the most direct purchase logic—“I just want to see what the homes you’ve installed look like”—and 沈亚楠 regards occupied case studies and user testimonials as key pillars of the brand architecture.
• At the time of recording, 栖息地 had roughly 1,600 users, including about 1,000 projects under construction and more than 100 users who had fully moved in. Some users already explain in comment sections the difference between Huawei-style post-installation smart systems and 栖息地’s integrated product. This shows that the core understanding can spread by word of mouth, though slowly.
• Customer acquisition ultimately rests on 3 pillars: brand appeal, repeat business from word of mouth and offline expansion. Brand is not traffic and cannot be directly attributed; it ultimately appears in visits and inquiries. Offline growth includes better store experiences, cross-industry partnerships and new-home development.

22. Creating a new category means building a ladder before claiming the first rung

• 沈亚楠 compares the categories: Tesla created the smart new-energy vehicle category, allowing Li Auto to choose a position on an existing mental ladder. 栖息地 must first prove that it is “not renovation,” then build a new ladder for integrated smart homes in the customer’s mind, and finally declare itself the creator and number one. The difficulty is an order of magnitude higher.
• Tesla itself once faced the old perception that electric cars were merely “toys for old men,” so it used the Roadster’s speed and smart display to change the category image. 沈亚楠 is looking for 栖息地’s equivalent emotional entry point: it cannot only explain why traditional renovation is bad; it must give the new category a vivid new picture.
• He believes a new category needs a “familiar stranger”: something too familiar fails to create interest, while something as alien as an extraterrestrial cannot be understood. 栖息地 must connect the familiar home with unfamiliar elements such as a complete product and smart chassis.
• Pop Mart, Pepsi, 元气森林 and 霸王茶姬 share the same pattern: they did not create demand from nothing, but captured social needs around IP, youth, zero-sugar health or inexpensive small pleasures. 栖息地’s technology and product trends are clear; it has not yet found an emotional narrative with the same penetrating power.

23. The era’s emotions behind home improvement may be longer stays, better aesthetics and cultural confidence

• Clear trends include people improving their current homes more seriously as moving frequency declines, Chinese users’ sensitivity to smart-home technology, and rising standards for aesthetics and daily life. But 沈亚楠 is looking for a more emotional trend and admits he “doesn’t have it yet,” even taking a vacation to sense real life more directly.
• In Dunhuang, he saw the scenic area make umbrellas, poles and even surveillance cameras match the colors of the desert and trees, so modern equipment would not disrupt the historical atmosphere. This convinced him that Chinese people are losing their tolerance for things that are “tacky and ugly,” and that visual and spatial aesthetics have become a genuine trend.
• The cultural confidence shown by drivers and guides in Dunhuang raises a second question: can 栖息地’s natural elegance connect to Chinese residential traditions rather than reproducing “foreigners’ beautiful living”? He is particularly interested in bringing the borrowed-view techniques of traditional gardens into modern homes.
• 沈亚楠 believes integrated smart homes may have to emerge in China first. American homes are larger and have more equipment rooms, reducing the need for integration; Chinese consumers place greater value on smart technology, creating a unique geographic advantage.

24. A RMB5T market can accommodate a trillion-yuan company, but near-term targets have been deliberately cut

• 沈亚楠 estimates that China has annual demand for roughly 20M new-home renovations and secondhand-home upgrades, representing a market of about RMB5T. The industry leader, 被窝, completes roughly 50,000 orders a year, yet no one can clearly name the second-largest player—evidence of extreme fragmentation.
• 栖息地’s end-state target is unchanged: take 20% of the market like Home Depot, equivalent to roughly RMB1T in revenue. Asked when it could get there, he first answered “10 years,” then said “about 15 years from now” after further questioning, calling the estimate relatively conservative and optimistic at the same time. JD.com, property developers, appliance companies, Haier, Midea, Huawei and Xiaomi could all enter through different paths.
• Near-term operations must face reality: the company is roughly 3 quarters behind expectations overall, lagging in product breadth, brand awareness, customer acquisition and delivery. The original annual bookings target in the 10,000s has been reduced to the 1,000s. Delivery itself is aimed at reliably meeting the roughly 85-day expectation of users, since some signed customers are still waiting for their homes to be handed over.
• The second complete product in the second half will not move sharply downmarket. 沈亚楠 believes that once layers of channels, duplicated customer acquisition and inefficient design delivery are removed, a customer spending RMB500,000 should not receive only RMB300,000 of value. Once scale, logistics, training and design tools mature, a good home he approves of might be built for roughly RMB250,000.

25. The $500M investment and end-state target ultimately point to an R&D platform and open ecosystem

• 沈亚楠 says he has continuously sold shares in Li Auto—“sell and invest, sell and invest”—and has personally invested roughly $500M. He also told his child: “There are only 2 things in the world that make you poor: gambling and entrepreneurship. So Dad is starting a company; you’ll have to work hard yourself.”
• He estimates that revenue of roughly RMB2B would allow the company to “start making money,” meaning cash flow would be sufficient to support continuous R&D. This is not static break-even: a technology company can become profitable quickly by cutting R&D, but then it is no longer a technology company. Current R&D remains “highly focused and slow-steady.”
• 栖息地’s R&D roadmap includes higher integration through cabinet-appliance systems, smart systems that know what is inside a cabinet and can help users find it, and household service robots. 沈亚楠 is willing to wait for embodied-intelligence companies to solve the “actuators and cerebellum,” while 栖息地 solves the “scenarios and brain.” He wants to see the product he imagines no later than 2028.
• Looking back on 3 years, he admits to 2 naive assumptions: overestimating the baseline quality of home-renovation workers and assuming that a good product would naturally earn a brand. Li Auto had 李想’s influence and successful track record as “air”; 栖息地 does not. The worst risk is not that the product is unreasonable, but that path dependence—like a keyboard layout—means something “reasonable” is not accepted within the timeline the founder can afford.
• The latest correction is a turn toward a more open, infinite-game mindset: the vision and mission are fixed, but direct operation, property-company partnerships and other methods can change. Renovation companies, studios, construction crews and third-party applications are no longer enemies. “A tree can never become too large just by growing taller.” Reaching trillion-yuan scale will ultimately require an ecosystem in which the product, logistics chassis and countless operators grow together.