Over the past two years, the question investors have asked Dingdong Maicai most often is: how are you different from Missfresh, and why did Missfresh fail while you survived?
In November 2023, Missfresh, once a high-profile startup, was delisted from Nasdaq. Around the same time, Wang Song, who had worked at Hema (known internationally as Freshippo) and Ele.me, became Dingdong Maicai’s CFO. His answer then was simple:
“We are a company that works on products, step by step.”
Back in 2021, before Missfresh’s sudden collapse, Dingdong founder and CEO Liang Changlin had already set a new course: efficiency first, scale later. He shifted from nationwide expansion to a region-focused strategy.
The following year, Dingdong doubled down. It cut losses and withdrew from Tianjin, Xiamen, Chengdu, and other cities, eventually exiting southwestern China entirely. In January 2024, the retrenchment peaked with the closure of 27 stations in Guangzhou and 11 in Shenzhen. Analysts doubted the company’s prospects.
Few expected that in the same year, Dingdong would halt its losses, return to growth, and post its strongest financials since going public. Revenue in 2024 rose 15.5% to RMB 23.066 billion (USD 3.2 billion), with net profit attributable to shareholders of RMB 295 million (USD 41.3 million). Operating cash flow reached RMB 929 million (USD 130.1 million) and turned positive for the first full year.
Just as Dingdong broke free from the cycle of trading scale for profitability, China’s retail market shifted again, with internet giants launching block-by-block fights for dominance in instant retail. For a company “neither too big nor too small,” the challenge is sustaining momentum and proving resilience.
Cutting losses and boosting efficiency “like Elon Musk building rockets”
The pandemic stress-tested the market, but it also fueled disorderly overexpansion that hastened the sector’s collapse. In July 2022, Missfresh abruptly disbanded. Suppliers across China scrambled for repayments, and many small merchants were left unpaid. With overlapping supplier networks, the fresh e-commerce sector quickly slid into a supply chain crisis.
Wang said that, at the time, Dingdong moved quickly to shoulder suppliers’ anxieties with absolute candor:
“We practically mobilized the whole company to give suppliers precise repayment schedules and accelerate payouts. After about one or two months, production stabilized again.”
One supplier told 36Kr that when Dingdong struggled to raise financing early on, Liang sold his own startup to sustain the company. “He had run a mother-and-baby e-commerce platform that made more than RMB 20 million (USD 2.8 million) in annual profit. We were moved when he shared this at a supplier gathering. We believed in his determination to make Dingdong work.”
Even so, Dingdong still had to prove it could generate cash on its own. From 2021–2023, it remained in the red.
Fresh e-commerce’s frontend warehouse model has clear drawbacks: limited SKUs, low basket sizes, thin margins, and high fixed costs—from spoilage to marketing and last-mile delivery. Competitors like Xiaoxiang Supermarket and Pupu Supermarket operate larger warehouses of 600–1,000 square meters with 4,000–8,000 SKUs, comparable to a midsize supermarket.
Early on, Dingdong’s sites, like Missfresh’s, measured about 300–400 square meters and carried 3,000–4,000 SKUs. Their smaller scale left less room for error and made operations more vulnerable to demand swings. After Missfresh collapsed, questions arose over whether Dingdong could survive.
Compared with rivals, Dingdong sensed the capital market shift earlier. From 2022–2023, it closed underperforming sites, cut losses, and stemmed the bleed. In May 2022, it exited parts of Hebei and Anhui; a year later, it shut all stations in Chongqing and Chengdu. Retrenchment reduced losses, while efficiency gains came from a deeper push into digitalization.
After becoming CTO in October 2020, Jiang Xu spent nearly a year upgrading Dingdong’s digital infrastructure so it could handle volume like a mature e-commerce platform. His team then focused on capturing daily operational data.
Jiang sees frontend warehouses as a distributed cold chain logistics system and wanted every link online. “If you assume 1,000 frontend warehouses and 4,000 SKUs per warehouse, the product combinations alone reach four million. That’s beyond human capacity,” he said.
The rollout was bumpy. Some days, warehouses received hundreds of boxes of milk far beyond sales averages. Other days, items sold out too early and there was nothing left to sell. “We sometimes joked that building our digital system was like Elon Musk building rockets, as it relied entirely on expansions to test and learn.”
Each inventory expansion prompted a root cause analysis, which over time reduced waste. After three years of iteration, Dingdong’s end-to-end digital system went live in late 2023.
Here’s how it works: every 24 hours, Dingdong forecasts SKUs for all warehouses. Factoring in staffing and inventory at central and frontend sites, it produces detailed production and replenishment plans. Suppliers receive purchase orders, then ship to central warehouses via Dingdong’s transport network. Some goods are processed or packaged centrally before allocation to frontend warehouses.
When consumers place orders, frontend warehouses pick and pack, and the system dispatches delivery capacity. Once fulfilled, orders feed back into the forecasting engine, closing the loop.
Gong Yunhao, Dingdong’s supply chain lead, focuses on warehouse-to-warehouse flows, especially transfers and management from central to frontend warehouses. To make a 500-square-meter site perform like a 1,000-square-meter one, he worked to speed up inventory transfers and make systems interlock more tightly:
“When goods reach the central warehouse, we arrange outbound and transfers. It’s critical to balance stockouts and spoilage. After allocating from central to frontend warehouses, we still have to ensure demand matching. If we see misalignment, we rapidly recall and reallocate products.”
Once the system was in place, Jiang began upgrading its depth of reasoning, gradually letting the system assume decisions previously made by people.
By late 2023, it handled 90% of procurement directly. Even in extreme weather, forecasting accuracy stayed above 85%. Loss rates, a key measure of operational efficiency, have held at 1–2% over the long term.
By then, the results of loss-cutting were finally showing.
2024: Focus, profitability, and becoming best-in-class in Shanghai
In September 2023, Dingdong posted its first monthly profit, followed by a profitable fourth quarter. Still, many in the industry believed it had traded scale for profitability.
“An investor told me at the time that our profits were saved, not earned. I understood the doubt,” Wang said. That is why in 2024 he focused on reigniting growth, starting with keeping people steady.
“At the start of last year, we brought our heads from offices in south China and Beijing to Shanghai. One day they saw us close Xiamen, the next day Chengdu, and they feared they would be next. We had to put their minds at ease.” That meeting, he said, became the turning point out of the loss cycle.
According to 36Kr, Dingdong cut losses by 60% in Beijing, Guangzhou, and Shenzhen in 2024, then redirected the savings to its strongest region: eastern China.
By then, fresh e-commerce had moved past its life-or-death moment. In August 2024, founder Liang announced a renewed expansion push, increasing warehouse density in Jiangsu, Zhejiang, and Shanghai.
From a broader retail perspective, China’s cold chain constraints, uneven supply chain coverage, and complex geography mean that—aside from Yonghui Superstores—no company has built a truly national brick-and-mortar grocery chain serving consumers’ “three meals a day.” The hypermarket model had been in decline even before retailers began adopting Pangdonglai’s playbook. Hema Fresh’s nationwide push into lower-tier markets repeatedly ran into food safety and supply chain issues.
Meanwhile, Fuzhou-based Pupu Supermarket stayed concentrated in nine cities, mostly in southern China, and in 2024 reported RMB 30 billion (USD 4.2 billion) in revenue and its first annual profit.
Wang noted that even during its toughest years, 2022 and 2023, Dingdong kept densifying its network in Jiangsu, Zhejiang, and Shanghai. In 2024, it added 130 sites in the region, focusing on second- and third-tier cities. Of the 25 cities where Dingdong operates frontend warehouses today, 15 are in Jiangsu and Zhejiang. Beyond first-tier cities such as Hangzhou, Nanjing, and Suzhou, it has entered lower-tier markets like Shaoxing, Wenzhou, Huzhou, and Taizhou.
The reasoning is straightforward: compared with unfamiliar northern markets, Dingdong understands consumer habits in Jiangsu and Zhejiang. In Wang’s view, disposable incomes in second- and third-tier cities are more stable than in first-tier cities, where consumption is more stratified.
Jiang Xu, Dingdong’s CTO, said new sites fall into two categories: warehouses created by splitting an existing site into two or three—sometimes using a hub-and-spoke setup—and entirely new locations that extend into fresh territory.
Wang added that Dingdong tunes warehouse density dynamically based on average orders per site over recent months and household coverage growth. The average delivery radius is 1.2–1.5 kilometers in Shanghai and about two kilometers in Jiangsu and Zhejiang.
“In the past, everyone said frontend warehouses were heavy investments with high costs,” Wang said. “But when you have 300 warehouses in Shanghai and each handles more than 1,500 orders a day, fulfillment costs drop exponentially.”
Wang believes Dingdong has reached the industry’s best levels for warehouse density, per-site order volume, and delivery cost in Shanghai.
The financials back this up: in Q4 2024, the fulfillment expense ratio fell 1.8 percentage points year-on-year (YoY) to 21.7%. For the full year, Dingdong operated more than 1,000 frontend warehouses, with average daily orders approaching 1,000 nationwide and 1,500 in Shanghai. The national average basket size exceeded RMB 70 (USD 9.8), and new sites typically broke even within three to six months.
Some locations have also been upgraded. In Shanghai, the average footprint of frontend warehouses has reached 500 square meters. Wang noted that roughly one-quarter of sites face lease expirations or upgrades each year.
“Compared with rejigging the network or expanding storage space, satisfying users is the harder task for us.”
Going beyond “three meals a day”
Some argue that Dingdong’s rapid digital rollout and path to profitability were aided by its small-warehouse model. But small sites also limit SKU breadth.
Wang said Dingdong runs a fast-on, fast-off selection mechanism: fresh categories have a 14-day assessment period, packaged goods 30 days. If sales miss expectations, items are removed immediately.
“We haven’t concluded whether 4,000 SKUs is optimal, or 2,000. We adjust SKUs for fruit, vegetables, and general merchandise quarterly,” he said.
A 36Kr comparison of product assortments across the mini programs of Dingdong Maicai, Hema Fresh, Xiaoxiang Supermarket, and Pupu Supermarket found few differences in SKU counts or prices for seasonal bestsellers like peaches, though Xiaoxiang offered the broadest range. In categories such as chilled fresh milk, fresh eggs, flowers and plants, sandwiches, and kitchen paper, large-warehouse operators Xiaoxiang and Pupu carried greater variety and higher counts.
Past media reports put Xiaoxiang’s SKU count above 10,000, approaching traditional supermarket scale.
By comparison, Hema’s assortment spans more categories but with fewer SKUs per category after a 2023 reduction. Dingdong’s focus on three meals a day gives it deeper fresh selections but fewer options for snacks, beverages, and household goods. Some consumers told 36Kr they could not find coffee beans or pet supplies on Dingdong, and that snack and juice choices were limited.
In 2024, Dingdong began expanding into leisure and snacking needs, moving beyond its three-meals core. Even so, some industry observers say category expansion remains too slow. One longtime fresh e-commerce practitioner said Dingdong’s higher basket size and profitability stem from two factors:
- Absorbing traffic from mom-and-pop stores and traditional markets, and gaining customers in Shanghai that Hema lost during its retrenchment.
- Adding higher-margin categories, particularly items less convenient to make at home, such as bakery goods and cooked meats.
That person also argued that Dingdong should treat fresh groceries as a traffic gateway and gradually build a broader, hypermarket-style assortment.
Private labels, often used to improve margins, now account for 35% of Dingdong’s sales, supported by in-house brands launched in recent years, such as Tsai Chang Ching.
“Dingdong wants its products to cover areas where suppliers have no edge or cannot meet our standards. It’s not just about tweaking margins,” Jiang said. He noted that the company is willing to take harder paths, such as building its own digital system or sourcing food directly, when it believes they are the right ones.
“For example, our chive pockets always use chives we grow ourselves, because chives and long beans are high-risk for pesticide residue,” Jiang said. Dingdong uses pest control and soil management, among other practices, to reduce residue as much as possible.
Internally, Dingdong refers to its “tap water philosophy,” Liang’s vision that high-quality ingredients should flow into homes as easily as tap water. After 2025, Liang launched a “one inch wide, one kilometer deep” strategy focused on good users, good products, good service, and good mindshare.
Wang added that for Dingdong, “good products” means long-term value and differentiation. The goal is to win repeat purchases through quality, not by chasing SKU count or undercutting prices on items with marketing-heavy labels.
To strengthen category expertise, Dingdong reorganized its product development center into ten independent business units, each led by a core executive. Wang oversees fruit, Jiang leads vegetables, and Gong Yunhao heads dairy, liquor, and beverages. The structure spreads top management’s time across building and refining product competence.
How will Dingdong Maicai respond?
Accolade Wines’ China general manager, Su Huaici, observed that after the 2025 Lunar New Year, Dingdong tightened quality control:
“Station managers manually check every bottle of wine, and any minor flaw gets it returned.”
But the competitive landscape has shifted. According to Su, partnering with instant retail platforms has become a trend for traditional wine companies, and the scramble for instant retail traffic now resembles the boom years of fresh e-commerce.
Instant retail, also called quick commerce, means ordering online and receiving goods within 30 minutes offline. Meituan’s Shangou warehouses illustrate the model: by not optimizing for in-store traffic, they can carry more SKUs than similarly sized frontend warehouses and face minimal spoilage.
As the instant retail space heats up, fresh e-commerce has entered a new round of price wars. In June, 7Fresh, previously focused on large hybrid facilities, announced 18 new locations in the Beijing-Tianjin region and pledged to widen its price advantage from 10% to 20%. Around the same time, Hema cut prices in Shanghai, lowering averages on more than 300 items by nearly 20%.
“In today’s environment, we have to be rational. Dingdong will not join a price war,” CFO Wang said. He outlined Dingdong’s plan to 36Kr: stick with the frontend warehouse model; deepen its presence in Jiangsu, Zhejiang, and Shanghai while keeping a foothold in Beijing; and keep upgrading algorithms and product capabilities.
Even while avoiding price wars, Dingdong stays active in instant retail through B2B partnerships with platforms such as Ele.me, Taobao Maicai, and Douyin. These linkages help keep it in the game.
Meanwhile, Xiaoxiang Supermarket has pledged to cover all first- and second-tier cities and expand into 200 quality county-level agricultural areas through direct sourcing. If Xiaoxiang secures markets outside eastern China, could Dingdong lose the chance to expand beyond its home base?
“In fresh e-commerce, we should be cautious about whether scale growth brings real economic benefit,” Wang said. “If the market expects Dingdong to keep growing 15–20% annually, eastern China’s growth alone can fully support that.”
He disclosed that Dingdong’s monthly household purchase penetration is about 30% in Shanghai and 25% in Suzhou and Hangzhou. “I’ve told investors many times: in eastern China, Dingdong still has at least twofold growth potential. Eastern China is not our final horizon, but when we expand will depend on supply chain capacity.”
As for why Dingdong survived while Missfresh failed, people most often cite steadiness and pragmatism, reflected in strong cohesion and fast reactions. At a supplier conference, founder Liang offered his view:
“In more than eight years of building this business, Dingdong has done only one thing: we bring vegetables from Xinjiang, Ningxia, Gansu, and Yunnan to Shanghai, Hangzhou, and Suzhou. The reason we’ve been profitable for ten straight quarters is also one thing: we reduced waste.”
“For more than three years, most of our decisions have been made by walking, watching, and adjusting as we go,” Wang added.
Wang said Dingdong values its quarterly meetings. Even if the core team has reached a conclusion, it still spends three days aligning everyone by discussing key questions, such as whether to roll out Shanghai’s product set to other cities, or whether to expand into new scenarios.
With the expansion narrative back under its control, Dingdong is now taking a more focused long-term path.
In Q1 2025, Dingdong’s net profit fell 34.96% YoY, raising questions about whether profitability had peaked. The company attributed the drop to short-term cost pressure from regional deepening and projected that both scale and margins would improve by year’s end.
In Liang’s plan, for the foreseeable future Dingdong will go deeper and be more focused on product strength, supply chain, and digital capability. Meanwhile, it will expand SKUs and its nationwide footprint more shallowly and slowly.
There is nothing new under the sun: a slower expansion may bring competitive pressure in the short term, but it helps a company adjust course in a volatile market. At Dingdong, that all stems from a collective rethinking of the old pursuit of scale for scale’s sake.
KrASIA Connection features translated and adapted content that was originally published by 36Kr. This article was written by Xie Yunzi for 36Kr.