11/16 2026
465

Produced by I Xiaohai fallsea
Written by I Hu Buzhi
In 2025, Roborock claimed the global No. 1 spot in robot vacuum cleaners with a 17.7% sales share, followed by Ecovacs at 14.3%—and the top five were all Chinese companies.
A decade ago, that position belonged to iRobot. In 2015, iRobot alone held 60% of the global market; by the third quarter of 2025, its share had shrunk to under 8%, and the company was acquired by the Chinese firm Chuancheng.
From iRobot's 60% to Chinese brands' nearly 70%, the market has changed hands. This is a textbook victory in going global, with Chinese brands leaving competitors far behind through supply chains and iteration speed.
The shift in market share took only ten years, but rebuilding profit may take longer. Speed itself is an overlooked cost. Chinese brands covered in ten years the road iRobot took twenty years to travel—but the price of speed is that profit has not yet caught up with scale.
Roborock's 2025 financial report, however, tells another story: revenue was 18.695 billion yuan, up 56.51% year on year, while net profit attributable to shareholders was 1.363 billion yuan, down 31.03% year on year.
Revenue grew 56%, but profit fell 31%. Selling more earned less. This company, ranked first globally in market share, earns an average net profit of only 235 yuan per machine sold.
Where did the money go? This is not just Roborock's problem, but a common question for the entire Chinese robot vacuum cleaner industry going global.
After ranking first in market share, profit did not follow market share. Instead, it was drained one account at a time by three costs. This paradox is the most real situation of the entire industry.
No one could have predicted that the one who won the market would instead lose profit. The most counterintuitive script in the business world is truly playing out in this category.
The narrative of going global usually involves both market share and profit rising, but robot vacuum cleaners have broken this rule. Economies of scale have failed here because channels have taken most of the incremental gains.
Scale did not bring bargaining power; instead, it brought deeper channel dependence. This is the underlying flaw of the going-global model in this category. The larger the scale, the deeper the dependence on channels, and the weaker the bargaining power. This is the underlying paradox of the going-global model in this category.
01 The Paradox of Ranking First in Market Share
In 2025, Roborock shipped 5.8 million robot vacuum cleaners globally, up 76.5% year on year. Its products entered more than 170 countries and regions and served more than 20 million households.
In Europe, during Prime Day 2026, Roborock captured 45% market share on Amazon; in North America, the figure was 33%, ranking first in both the United States and Canada.
In the first half of 2026, Roborock's overseas revenue reached 6.073 billion yuan, up 53.77% year on year, accounting for more than 60% of total revenue.
Of the 2.181 billion yuan increase in revenue that quarter, 2.123 billion yuan came from overseas, accounting for 97%. The domestic market is already close to saturation, and overseas is the only source of growth.
Roborock's choice was not to go global, but that it had to go global. Yet the growth did not turn into profit; instead, it turned into a heavier cost burden. The logic of going global has been rewritten here.
In the past, going global was about finding growth; now going global itself requires continuous investment, with growth and costs rising almost in tandem.
When growth and costs rise together, scale is no longer an advantage but a burden. Profit release must wait until the cost curve turns downward.
Before the cost curve turns downward, all scale growth is merely working for channels, and brands cannot capture the profit that should belong to them.
In 2025, Roborock's selling expenses reached 4.894 billion yuan, up 64.95% year on year, with a selling expense ratio of 26.18%. For every 100 yuan in revenue, the company spends 26 yuan selling goods.
Within selling expenses, advertising and market promotion expenses were 3.259 billion yuan, up 69.42% year on year, and platform service fees and commissions were 868 million yuan. The two together accounted for 84% of selling expenses.
This 4.127 billion yuan is three times Roborock's full-year net profit of 1.363 billion yuan. For every yuan Roborock earns, it must first give three yuan to platforms and advertisers.
This is not a problem of operating efficiency, but of channel structure. Money flows directly from the brand's pocket into the platform's pocket.
The money was not earned by suppliers, nor swallowed by logistics; it was collected by channels. Amazon's commissions, Google's ads, and entry fees for offline stores—every one is a toll.
Channels are not a cost center but a profit black hole. When an industry's selling expense ratio exceeds 25%, brands have already lost pricing power.
Brands that lose pricing power can only keep increasing investment, falling into a vicious cycle of selling more and losing more, and losing more yet having to sell more, unable to extricate themselves.
This is why ranking first in market share and ranking first in profit separate. Scale brings not bargaining power, but deeper channel dependence—the more you sell, the less you can do without platforms.
Once a channel siege is formed, it is hard to break, because the cost of exit is higher than the cost of continued investment, leaving no choice but to push forward.
The real price of ranking first in market share is turning profit from product price differences into channel tolls.
02 The First Account: Platform Commissions and Advertising
Amazon is Roborock's largest overseas traffic gateway. Behind the 45% Prime Day share in Europe and 33% in North America is real money.
Amazon charges 8% to 15% commission on large items such as robot vacuum cleaners, plus FBA delivery fees. For a machine priced at 500 US dollars, platform commissions alone may exceed 60 US dollars.
The bigger expense is advertising. Amazon's ad bidding model means that the more sellers there are, the higher keyword prices become. As a latecomer, Roborock must keep raising bids to rank at the top.
Advertising is not a cost but a ticket. In Amazon's ecosystem, no ads means no traffic, and no traffic means no sales. This is a cycle that cannot be escaped.
The essence of ad bidding is a zero-sum game among sellers, with the platform profiting while brands consume each other but none dares to stop first. The platform controls the traffic switch, and sellers can only passively follow price increases.
Under this model, the money brands earn ultimately becomes platform advertising fees. In 2025, Roborock's advertising and market promotion expenses grew 69.42%, far exceeding revenue growth of 56.51%. Advertising spending ran ahead of revenue, showing that Roborock is using money to buy share.
Trading profit for share is a dangerous game. Once ad spending slows, share may immediately decline, meaning Roborock simply does not dare to stop.
Once the game of trading profit for share starts, it is hard to stop. Stopping means all previous investment goes down the drain.
This is not a problem unique to Roborock. Ecovacs' 2025 selling expenses were 5.938 billion yuan, with a selling expense ratio of 31.19%, including 4.741 billion yuan in advertising, marketing, and platform service fees.
The two companies together spend nearly 9 billion yuan a year on advertising and platform commissions, a figure that exceeds the full-year net profit of most Chinese home appliance companies.
But the results are completely different. Roborock's net profit fell 31%, while Ecovacs' net profit rose 118%. Why are the results so different when both are burning money? Burning money in different channels yields completely different returns. This is the best proof that channel choice determines the income statement.
The answer lies in channel structure. Roborock's direct sales revenue was 8.973 billion yuan, and distribution revenue was 9.708 billion yuan. Online channels account for the bulk, especially Amazon.
Ecovacs' approach is different. In the first half of 2026, overseas offline revenue grew 47.1% year on year, US Amazon grew 10.7%, and German Amazon grew 13.1%. Offline growth was three to four times that of online. Ecovacs entered Walmart and Costco in the United States, and high-end stores such as Fujisaki and Mitsukoshi in Japan.
These channels have high entry fees, but traffic costs are far lower than Amazon bidding. Once inside, users come for the store, not for the ads.
Offline stores bring their own traffic. Brands only need to pay entry fees and do not need to pay separately for every click. Roborock is also expanding offline, having entered Costco, MediaMarkt, Carrefour, and Yamada Denki, but its base remains online, with self-operated channels accounting for 65% in Europe.
Amazon is the core engine of scale growth and also the main outlet for profit loss. Roborock has yet to find a solution to this contradiction. Amazon is Roborock's growth engine and also its profit funnel, and there are no signs of this contradiction being completely resolved in the short term.
The value of offline channels is being rediscovered. They are not the enemy of scale but the friend of profit. The key is whether one is willing to trade short-term growth for long-term profit.
There is no standard answer to this choice. If you want scale, accept low profit; if you want profit, accept slower growth. So far, no one in the robot vacuum cleaner going-global business has been able to achieve both at the same time.
Channels determine the income statement, not products. Amazon gives you scale, offline gives you profit, and high-end gives you premium.
There is no standard answer to the trade-off between scale and profit, but every choice corresponds to different risks and returns. The key is to be clear about what you want.
03 Three Approaches, Three Kinds of Profit
The three leading companies took three different paths and delivered three completely different income statements. Behind the differences is the divergence in channel choices.
Roborock focuses on Amazon and DTC independent sites, following the scale route: revenue growth of 56.51%, global shipments of 5.8 million units, and first place in market share, but net profit down 31.03%.
Online advertising and commissions consume most of the profit. The larger the scale, the greater the loss. This is Roborock's most awkward current situation.
Roborock's dilemma is not an isolated case, but a problem faced by all Chinese sellers dependent on Amazon. The bigger the platform, the weaker the seller.
All sellers dependent on a single platform face the same dilemma, and Roborock is simply the most typical sample among them.
Ecovacs focuses on local subsidiaries and offline stores, following the profit route. Revenue growth is only one quarter of Roborock's, but net profit attributable to shareholders was 1.758 billion yuan, up 118.13% year on year.
Net profit excluding non-recurring items grew 126.14%. Ecovacs earned 400 million yuan more than Roborock but sold fewer goods. This is the power of offline channels.
The returns from offline channels are slow to heat up, but once established, they are hard to shake. This is the fundamental reason Ecovacs' profit far exceeds Roborock's.
Traffic costs are low, user repurchase rates are high, and brand premium capability is strong. Offline stores are a slow business, but once you gain a foothold, profits are much thicker than online.
Offline is a slow business that requires time to accumulate and patient investment, but once you gain a foothold, the thickness of profit far exceeds online.
Dreame took the third path: high-end. In the first quarter of 2026, Dreame took global first place in both unit share at 23.7% and sales share at 28.0%.
More than 70% of Dreame's revenue comes from overseas. In the high-end model market above 700 euros in Northeast Europe, its market share reached 70%, and in North America it grew 185% year on year.
High-end means high average selling price, and high average selling price means gross margin has more room to cover channel costs. For a machine selling at 1,000 euros, even if channels take 30%, the profit left for the brand is greater than for a model selling at 300 euros. The three approaches have no right or wrong, but they have costs. Roborock gained scale but lost profit, Ecovacs gained profit but lost growth, and Dreame gained high-end positioning but still has a small scale.
Whether these three paths can succeed depends on one premise: whether channel costs can be reduced. If Amazon ad bidding continues to rise, offline entry fees remain high, and high-end competition intensifies, all three will face the same pressure.
Channel costs are the sword of Damocles hanging over all going-global brands, and no company can truly stay outside it. In 2025, Roborock's selling expense ratio was 26.18%, and Ecovacs' was 31.19%. Both spend 26 to 31 yuan out of every 100 yuan in revenue selling goods, while the white goods industry usually spends only 10% to 15%.
The channel cost of robot vacuum cleaners is two to three times that of white goods. This price difference is the entry fee for ranking first in market share. The choices of the three companies are essentially different rankings of "scale" and "profit." There is no right or wrong, only who can endure more.
In the end, this battle is not about speed but endurance. Whoever can hold on until channel costs fall will laugh last.
04 The Second Account: Logistics and Overseas Warehouses
Beyond channels, the second account is logistics and overseas warehouses. Roborock's overseas gross margin in 2025 was 48.15%, down 5.55 percentage points year on year.
5.55 percentage points may not sound like much, but on overseas revenue of 10.442 billion yuan, it means 580 million yuan in profit evaporated, nearly half of Roborock's full-year net profit.
The decline in gross margin was not driven by product price cuts but by rising costs. In Europe, Roborock's self-operated channels account for 65% of sales, which means the company must shoulder warehousing, logistics, delivery, and after-sales service itself.
Self-operation means bearing all the risks yourself: when sales are strong, you earn more; when they falter, you lose more. Self-operated channels sound attractive—control over end-user data, no reliance on distributors, and ample profit potential—but they rest on heavy investment.
Overseas warehouses, local teams, and after-sales networks are all fixed costs, and these costs do not decline in proportion to sales growth.
The defining feature of fixed costs is that they must be paid regardless of how much is sold, which means the self-operated model inherently requires high sales volumes to dilute costs. Roborock has established local teams in the United States, Germany, and South Korea, gradually localizing marketing, sales, logistics, warehousing, and after-sales service. The labor costs of these teams are several times those in China.
The rent and operating costs of overseas warehouses are also considerable. Shipping a machine from China to an overseas warehouse in Europe costs about 30 to 50 US dollars in sea freight; warehousing runs about 3 to 5 US dollars per month; and last-mile delivery costs about 15 to 25 US dollars.
Under the direct mail model, these costs can be passed on to consumers, but under the self-operated model they can only be absorbed by the company. Overseas warehouses weigh on not only logistics costs but also inventory.
Under the self-operated model, inventory is another form of cash. Goods sitting in warehouses incur costs and depreciation every day.
Unsold goods lying in European warehouses generate costs every day. Each day of slower inventory turnover shaves a little more off profit. Overseas warehouses are a double-edged sword: used well, they can improve delivery speed and user experience; used poorly, they become an inventory black hole that continuously burns money.
Used well, overseas warehouses are a moat; used poorly, they are a bottomless pit for burning money. The key lies in inventory turnover efficiency. Ecovacs' approach is local subsidiaries plus distribution networks, sharing warehousing and logistics pressure through local partners. Although it gives up part of the profit, it also reduces the heavy-asset risk of self-operation.
This is why Ecovacs has a higher selling expense ratio yet net profit has surged instead. Selling expenses are high because it gives profits to distributors, but in exchange it gets lower inventory risk and more stable cash flow.
The distribution model sacrifices part of the profit but gains risk diversification and cash flow stability. This is a clear-headed trade-off.
The story of asset-light going global no longer works in the robot vacuum cleaner category. Overseas warehouses, local teams, and after-sales networks are all heavy assets.
Heavy assets are not a bad thing. They are both a moat and a threshold. The key is whether sufficient scale can dilute fixed costs to a reasonable level.
05 The Third Ledger: Compliance and Geopolitics
On July 28, 2026, the U.S. Federal Communications Commission issued Public Notice DA 26-786, adding foreign-produced advanced robotic devices to the covered list, and robot vacuums were prominently included.
Devices on the list are in principle unable to obtain new FCC equipment authorization and cannot be imported, marketed, or sold in the United States. The criteria are clear: autonomous mobility, weight over 2 kilograms, and environmental perception and networking capabilities.
The good news is that existing models that have already obtained FCC certification are unaffected. The bad news is that the robot vacuum industry iterates two to three generations a year, and the inability of new products to obtain new authorization means the iteration rhythm is interrupted. Top brands Roborock and Ecovacs have relatively complete certification reserves, giving them a relative advantage in the short term, but the threshold for small and medium-sized brands and new entrants has been substantially raised.
The competitive landscape in the U.S. market will further concentrate toward the top, and the FCC has inadvertently become a supply-side reform driver for China's robot vacuum industry. FCC regulation has unintentionally helped top brands clear out small and medium-sized competitors, and market concentration will further tilt toward the leaders.
But the benefit is short-term. In the long run, compliance costs will only get higher, not lower, and the threshold will keep rising.
The FCC is only part of compliance costs; tariffs are the more direct burden. The United States imposes a 25% Section 301 tariff on Chinese robot vacuums, and starting July 24, 2026, an additional 12.5% forced labor tariff was added, totaling 37.5%.
For a robot vacuum with a landed price of $200, tariffs alone amount to $75. Add Amazon commissions, FBA fees, and advertising spending, and the profit margin left for brands is squeezed to the extreme.
These costs are not one-time but ongoing. Tariffs must be paid every year, FCC certification must be done for every new product, and data compliance has different requirements in every market. The characteristic of compliance costs is that they only increase, never decrease. Every policy tightening means new investment and new thresholds.
Europe has GDPR, the United States has the FCC and state privacy laws, and Southeast Asian countries have inconsistent certification standards. Every time a new market is entered, a new set of compliance must be built, and every set of compliance costs real money.
Compliance costs are not one-time fines but a set of gates that continuously collect tolls. Certification, tariffs, data compliance—each gate passed thins the profit by another layer.
06 Penetration and Southeast Asia: Where Is the Growth
Since profits are so hard to earn, why do Chinese brands continue to burn money? Because penetration is still low and market space is still large.
The penetration rate of robot vacuums in Chinese households is 5.7%, and in the United States it is 20.5%. By comparison, the penetration rate of vacuum cleaners in European and American households exceeds 90%.
Robot vacuums have not even reached a fraction of vacuum cleaners, which means the market is far from its ceiling, and money to educate users still has to keep burning.
Low penetration means large market space, but it also means high user education costs, and this ledger cannot be balanced in the short term. In 2025, global robot vacuum shipments were 24.124 million units, up 17.1% year on year, and global cleaning robot shipments were 32.72 million units, up 20.1% year on year.
The Middle East and Africa market grew 95.6%, and Central and Eastern Europe grew 40.3%. These numbers mean robot vacuums are still in the early stage of penetration, and growth mainly comes from emerging markets rather than Europe and the United States.
Emerging markets are growing quickly, but average selling prices and profit margins are far lower than in mature markets, making it difficult to achieve both volume and quality.
Southeast Asia is the most typical growth market. The share of Chinese brands in Southeast Asia's robot vacuum market soared from 38.8% to 64.1% within a year, and Ecovacs ranked first in Vietnam with a 34% sales share.
In the first five months of 2026, China's cleaning robot exports reached 14 billion yuan. Southeast Asia, the Middle East, and Latin America—markets that were neglected in the past—are becoming new battlegrounds for Chinese brands. But emerging markets are not a free lunch. TikTok Shop's platform commission in Southeast Asia ranges from 5% to 21% depending on market and category, transaction fees are 2% to 6%, and with advertising spending added, total platform deductions can reach 28% to 37%.
In other words, selling a robot vacuum in Southeast Asia means the platform takes nearly one-third. Compared with Europe and the United States, Southeast Asia has lower average selling prices but a higher share of channel costs. Emerging markets are not a profit haven; the logic of channels is the same in any market, only expressed in a different form.
Emerging markets are not a profit haven but another version of the channel siege. The market has changed, but the logic of channels has not.
After reaching first place in share, it is not a harvest period but the starting point for larger-scale investment. The penetration ceiling is still far away, and the battle in Southeast Asia has only just begun.
Emerging markets are not a safe harbor. The logic of channel commissions holds in any market; it is merely a different set of platforms and rules.
07 The Real Test After Reaching First Place in Share
Roborock sold the most globally but its profit margin is lower than Ecovacs'; Ecovacs has deeper offline coverage but its growth rate is lower than Roborock's; Dreame is aggressive in high-end segments but still small in scale.
Each of the three companies has its own ledger and its own difficulties. Roborock needs to solve the profit problem, Ecovacs needs to solve the growth problem, and Dreame needs to solve the scale problem. The three companies' problems are different, but the underlying logic is the same: it is a game between channel costs and profits.
But one problem is common: whoever can lower the channel tolls can truly turn share into profit. There are three ways to reduce channel costs. The first is to reduce dependence on a single platform, develop DTC independent sites and offline channels, and lower the share of Amazon commissions and advertising.
Roborock's self-operated channels in Europe account for 65%, which is exactly moving in this direction, using self-operation to dilute platform commissions.
Self-operation is a long road that requires sustained investment and patience, but once it works, it can escape deep dependence on platforms. The second is to use brand premiums to cover channel costs. Dreame holds 70% of the high-end market above 700 euros in Northeast Europe, and high average selling prices naturally lower the share of channel costs.
The third is to use scale to dilute fixed costs. Overseas warehouses, local teams, and compliance systems are all fixed investments. The more units sold, the lower the cost allocated to each machine.
The premise of economies of scale is that fixed costs are large enough and variable, and overseas warehouses and local teams fit exactly this characteristic. Roborock's shipment volume of 5.8 million units is approaching this critical point, but the release of economies of scale takes time and requires sustained investment. The critical point has not yet been reached, and Roborock still needs to continue investing; profit release may have to wait until the next operating cycle.
Economies of scale do not arrive automatically. They require channel structure, inventory management, and local operations to be in place at the same time; missing any one link will slow the release of profits. Economies of scale do not arrive automatically; they require simultaneous improvement in operational efficiency, and any weak link will slow the release of profits.
Before channel costs come down, first place in share is just an admission ticket that requires continuous payment. iRobot took twenty years to fall from 60% to less than 8%, while Chinese brands took ten years to go from 0 to nearly 70%.
In the next ten years, the test is not whether they can defend their share, but whether they can turn share into profit. There is no standard answer to this question, only bill after bill to be paid. First place in share is only an admission ticket, not a pass. The real test begins the moment first place is achieved.
First place in share is not the finish line but an admission ticket that requires continuous payment. Whoever can lower the channel tolls can truly turn share into profit.