“Local services” hides two ledgers. To-home is food delivery, flash marts, instant retail—money follows riders; density and subsidies set per-order math. In-store is group-buy, travel, and local leisure—money follows redemptions and ads; fulfillment is light and margins thick, but traffic can be skimmed by content apps. Meituan runs both books; DoorDash is mostly to-home logistics; Grab layers mobility and fintech beside deliveries. The profit-pool gap starts with one question: are you selling that last kilometer of delivery?

To-home: a high take rate is not a thick profit

To-home costs are physical: match/tech fees plus fulfillment fees, rider incentives, insurance, support. Cross-border comps often note higher overseas net monetization (ex-delivery) with consumers carrying more of the fulfillment premium, while China more often blends subsidies, merchant take, and delivery fees into price wars. Result: overseas take rates can look prettier without unit profits clearly beating Chinese peers—operating leverage and “willingness to wage war” cut sharper at home.

DoorDash’s filings frame the goal as optimizing absolute profit dollars, not a sacred take-rate percentage. Q1 2026 materials put Net Revenue Margin near the low-13% band; Q2 kept Marketplace GOV and Adjusted EBITDA growing, with Adjusted EBITDA around ~2.8% of Marketplace GOV (~2.4% in Q1). Teardowns stress gross margin can expand while take rate barely moves—via density, batching, ads, and fulfillment efficiency, not endless commission hikes. The to-home pool is fulfillment curves + ad attach; take rate is only the door.

Grab’s Deliveries segment writes “thin” plainly: FY2025 Deliveries Adjusted EBITDA margin on GMV about 2.0%, Q4 about 2.2%; company remarks for Q2 2026 put that margin near 2.3%. Growth rides GMV and advertising; profit rides leverage and ads—not Meituan-style in-store OPM cash cows. Deliveries are Grab’s scale and frequency engine; historically thicker segment profits sit more in Mobility and peers. To-home local commerce itself remains single-digit GMV-margin grind.

Meituan’s to-home ate a subsidy war in 2025: about ¥23.4B full-year net loss, sales and marketing roughly ¥102.9B, core local commerce flipping to a loss; annual reads still put food delivery above ~60% GTV share while delivery-service revenue felt the fight. Q1 2026 company release: core local commerce revenue about ¥64.1B, operating loss narrowed to about ¥2.0B from about ¥10.0B prior quarter as “anti-involution” and rival pullback cooled subsidies. Public bridges often show food delivery still soft YoY, flash purchase faster, in-store/travel middling—to-home heals through UE, not overnight high-margin in-store magic.

In-store: light fulfillment, fear of traffic detours

In-store does not scale a rider army one-for-one with orders. Historical research long cast Meituan in-store/travel as high-gross, commission plus online marketing—OPM clearly thicker than delivery; some models centered in-store OPM near the low-**30%**s (year assumptions differ; direction is “thicker than to-home”). Recent commentary still treated in-store as a stable profit pillar—until Douyin poured resources into in-store group-buy.

Competition flips: to-home fights instant logistics density; in-store fights who seats the guest, who redeems, who sells merchant ads. Content seeding can inflate GMV fast while redemption lags shelf-search modes—impulse tickets, unredeemed fat. Broker notes echoed in industry copy: traffic edge is real; redemption gap is real. Meituan keeps the pool but spends to defend share, compressing OPM; content platforms can treat in-store as a merchant-budget battlefield and prioritize scale over profit for a while.

Mismatch: delivery wars burn Meituan cash; Douyin’s in-store push burns Meituan’s profit foundation. To-home can repair UE when regulators cool subsidies; once in-store becomes a pure distribution contest, moats need reviews, consumer protection, and SMB tools—not denser short video alone.

Three platforms: which layer holds the pool

Platform	To-home role	In-store / other thick layer	Profit intuition
Meituan	Core moat; subsidy eras can swallow group profit	In-store/travel was cash cow; now content-skimmed	Two ledgers cross-subsidize; wars let to-home drag the story
DoorDash	Business is mostly the marketplace	Ads, membership, fulfillment lift absolute profit	Pool in density + ads, not a parallel “in-store division”
Grab	Deliveries for scale; ~2% GMV margins	Mobility thicker; fintech still investing	To-home is the door; group profit does not ride delivery alone

DoorDash lacks Meituan’s parallel high-margin in-store leg, so the Street reads Marketplace GOV, Net Revenue Margin, ads, and Dasher seasonality—not group-buy OPM. Grab’s super-app stacks deliveries, mobility, and payments; Deliveries’ ~2% margins can ride other segments and ads. Meituan’s uniqueness: thick in-store and heavy to-home in one app—and thus exposure to both flash/food price wars and Douyin in-store diversion.

Sources

• Meituan Q1 2026 results
• DoorDash Q1 2026 financial results
• Grab Q4 and FY2025 results