Delivery Apps Cut Rates 2%p, Raised Delivery Fees ₩500

Why the Same Question Remains After Fee Cuts

Reader, on February 26, 2025, Baedal Minjok (Baemin)—South Korea’s top food delivery platform—reduced its brokerage commission on “Baemin Baedal” bundled delivery orders from a flat 9.8% to a tiered rate between 2.0% and 7.8%, based on restaurant revenue. Coupang Eats adopted the identical tiered structure on April 1. Both rollouts were the direct outcome of an agreement reached in November 2024 by the government-backed Delivery Platform-Merchant Win-Win Consultative Body.

Yet the numbers that followed point in the opposite direction. In a survey released this September by the Korea Federation of SMEs (KBIZ), merchants spent an average of 26.2% of their monthly sales on platform transaction costs—up 5.2%p from the survey 1 year earlier. During the October National Assembly audit, lawmakers revealed calculations showing that on an order for a single pork cutlet priced at ₩18,000 (~$13.0), as much as ₩6,943 (38.6%) went straight to platform costs.

Why do numbers like this keep appearing even after commission rates were cut? In this issue, we trace that question line by line through actual settlement statements. We examine how advertised commission percentages translate into real financial burdens, the contractual clauses that barred merchants from setting different prices on other apps, and the gap between the celebrated win-win agreement and its real-world application.

---

Settlement statements deduct far more than just a brokerage fee

When a restaurant receives orders through platform-dispatched delivery services like Baemin Baedal¹ or Coupang Eats, its final settlement statement deducts roughly the following items:

Item	What it pays for	Relation to order value
Brokerage fee	Connecting and matching orders	Proportional to order value (2.0–7.8%)
Payment processing fee	Credit card and digital payment processing	Proportional to order value (~3% based on press reports)
Merchant-borne delivery fee	Store’s share of rider delivery costs	Fixed amount per order (₩1,900 (~$1.4) to ₩3,400)
Value-added tax (VAT)	Tax levied on the above fees	Proportional to total fees
Ad & coupon expenses	Exposure and discounts	Chosen by the store; often billed separately from settlements

What is worth noting here is the blend of two fundamentally different fee structures. Brokerage and payment processing fees are percentage-based, meaning they shrink when order values fall. Delivery fees, by contrast, are flat per-order charges that do not budge regardless of how small the ticket is. As a result, under the very same rate schedule, restaurants with lower average ticket sizes see their total fee burden rise sharply. This structural dynamic is precisely why Baemin itself disclosed in June 2025 that on a ₩10,000 order, the total burden borne by the merchant exceeded 40%.

In a survey by the Korea Federation of SMEs (KBIZ), breaking down the transaction costs of platform-listed merchants showed brokerage fees accounting for 42.0%, delivery fees 29.9%, advertising costs 15.6%, and payment processing fees 12.4%. Brokerage commissions accounted for less than half the total. That is why judging where total merchant costs have moved based on headline commission rates alone is misleading. The survey drew pushback from the Korea On-Line Shopping Association (KOSA), which argued that simply lumping together disparate costs exaggerated the perceived burden. That is why this issue tracks shifts item by item rather than focusing solely on the aggregate total.

The break-even point for the top 35% tier is ₩25,000

The win-win fee structure did not merely lower brokerage commissions; it also reset merchant-borne delivery fees across different revenue tiers. Placing the terms announced by Woowa Brothers (operator of Baemin) in January 2025 alongside the prior fee scheme (a 9.8% brokerage fee and delivery fees of ₩1,900 to ₩2,900) yields the following comparison:

Sales rank tier on Baemin	Brokerage fee	Merchant-borne delivery fee	Compared to previous
Top 35%	7.8%	₩2,400–₩3,400	Rate cut by 2.0%p, delivery fee raised by ₩500
Over 35% to 50%	6.8%	₩2,100–₩3,100	Rate cut by 3.0%p, delivery fee raised by ₩200
Over 50% to 80%	6.8%	₩1,900–₩2,900	Rate cut by 3.0%p, delivery fee frozen
Over 80% (Bottom 20%)	2.0%	₩1,900–₩2,900	Rate cut by 7.8%p, delivery fee frozen

Let us isolate the calculation for the top 35% tier. When the commission rate drops by 2.0%p, the brokerage savings per order equal 2% of the order value. In return, the delivery fee went up by ₩500 per order. The order value where the two offset each other is ₩25,000.

• ₩20,000 order: brokerage fee falls by ₩400, delivery fee rises by ₩500 → net increase of ₩100 per order.
• ₩25,000 order: brokerage fee falls by ₩500, delivery fee rises by ₩500 → zero net change.
• ₩30,000 order: brokerage fee falls by ₩600, delivery fee rises by ₩500 → net savings of ₩100 per order.

Payment processing and advertising fees were left out of this comparison because they did not change. While delivery fees vary within a set range based on region and distance, both the floor and the ceiling of the range rose by ₩500, leaving the net difference identical. Running the same math for the tier between 35% and 50% puts the break-even threshold at roughly ₩6,700, meaning costs drop on virtually every order.

This arithmetic matches the examples Baemin provided at the time of its announcement. Using an average order value of ₩25,000, Baemin explained that the bottom 20% tier would save ₩1,950 per order, while the two middle tiers would save ₩750 and ₩550, respectively. The top 35% tier was conspicuously missing from that breakdown. Under the math above, the net change for the top tier at that order size is precisely ₩0. Given that press reports in 2025 cited the average order value across food delivery apps at around ₩26,000, the average order for stores in the top tier sits right on the edge of breaking even.

The methodology used to assign tiers is equally critical. Baemin determines tiers based on daily average delivery sales over the preceding 3 months on Baemin 1 Plus (its bundled delivery service), whereas Coupang Eats uses actual monthly sales. Because classification relies on relative rankings within the app rather than absolute revenue figures, busier stores inevitably end up in the top tier. This explains why merchant associations pushed back vigorously just before implementation, warning that a large share of store owners, including franchise operators, would find themselves lumped into the top 35%.

When the KBIZ survey asked merchants about the cost-saving impact of the tiered fee structure, 49.4% answered “no change,” 35.4% said “unhelpful,” and 15.1% reported it was “helpful.” Because this survey was conducted across 350 platform-listed merchants as a whole and was not broken down by tier, it cannot be tied directly to the calculations above. Still, the fact that the tier with the smallest rate cut and the steepest delivery fee hike belongs to the highest-grossing stores goes a long way toward explaining why merchant sentiment remains so sharply divided.

When You Can’t Charge Different Prices on Other Apps

Another way to cushion the blow of rising commissions is menu pricing. If a restaurant raises prices slightly only on the app that hiked its fees, the customers on that specific platform end up sharing the burden of those higher costs. Blocking this workaround was precisely the allegation behind the Korea Fair Trade Commission’s (KFTC) investigation into Most-Favored-Nation (MFN)² clauses.

According to the KFTC’s findings, Coupang (starting in March 2023) and Baemin (starting in May 2024) restricted eligibility for their free delivery programs—Wow Store and Baemin Club, respectively—to restaurants that did not set food prices, minimum order amounts, or discount conditions at a disadvantage compared to rival apps. Stores that failed to comply were excluded from these benefits.

Here is where this condition ties directly into settlement statements: even if a platform hikes its commission, restaurants can’t simply raise their prices on that single app. The only options left are absorbing the costs or raising prices across all apps uniformly. Conversely, an app charging lower commissions cannot differentiate itself by signaling lower prices to consumers. This mechanism is why the KFTC, in issuing its examination report to both companies in October 2025, identified MFN clauses as the key driver stifling platform competition and pushing up commission rates. Based on order values disclosed by the KFTC, Coupang Eats’ market share surged from the 10% range in 2023 to the 30% range in 2025, while Baemin slipped from the 80% range to the 50% range. Together, the two still dominate virtually the entire market.

Here is how the case unfolded:

• October 13, 2025: The KFTC issued an examination report alleging violations including MFN demands and Coupang’s tying practices with its Wow Membership.
• Application for a consent decree: Coupang and Baemin applied for a consent decree³ and submitted win-win relief packages. Baemin offered a ₩300 billion (~$216 million) relief plan over 3 years and the abolition of MFN clauses. Coupang proposed ₩60 billion in merchant support over 4 years, the removal of MFN-related terms, and decoupling Wow Store status from free delivery perks.
• May 27 and June 10, 2026: The KFTC plenary session determined that neither application met the criteria for initiation, announcing their rejection on June 18.

The case proceeded to formal deliberation on the merits. The relevant sales⁴ tied to the alleged MFN demands were cited at approximately ₩730 billion for Baemin and ₩710 billion for Coupang. The case examiners estimated fines of roughly ₩239 billion to ₩510 billion across 3 Baemin cases, and roughly ₩25 billion to ₩42 billion for Coupang’s MFN case. The final penalty will be decided by the commission. Both platforms claimed they had already ended MFN demands, while the KFTC maintained that it still needed to verify whether they had actually ceased. As of November 1, no public reports of a final ruling on the merits could be found. The KFTC has stated it aims to reach a conclusion before the end of the year.

What Is That Same 7.8% Being Multiplied By?

Even with the exact same fee percentage, the payout changes if the base amount differs. In October 2025, the KFTC issued a corrective recommendation against a clause in Coupang Eats’ terms of service that calculated brokerage and payment fees based on the pre-discount price—even when the restaurant funded the discount entirely out of pocket.

The KFTC offered this example. If a customer pays ₩15,000 for a ₩20,000 dish after applying a merchant-funded ₩5,000 coupon, a 7.8% fee assessed on the discounted total comes out to ₩1,170; calculated on the pre-discount total, it jumps to ₩1,560. Measured against the ₩15,000 the store actually received, the effective fee rate becomes 10.4%. The KFTC noted that most food delivery platforms, including Baemin and Yogiyo, calculate fees based on the post-discount amount. When confronted during the National Assembly audit on October 23 about Coupang Eats not yet adopting the recommendation, its CEO responded to the effect that the matter was undergoing further internal review.

---

The Gap Between Win-Win Pledges and Settlement Statements

On June 19, 2025, Baemin announced an interim agreement reached through social dialogue with merchant associations: it would completely waive brokerage fees on orders of ₩10,000 (~$7.2) or less and provide tiered fee relief for orders between ₩10,000 and ₩15,000. The platform estimated the initiative at up to ₩300 billion over 3 years. It was an explicit response to criticism that the rise of single-portion dining placed a disproportionately heavy fee burden on small-ticket orders.

The scope of coverage was contentious from the start. While a Baemin official claimed that orders of ₩15,000 or less made up roughly 1/3 of total platform volume, a survey of 502 merchants conducted by the Consumer Public Interest Network revealed that the average minimum order threshold across Baemin storefronts was ₩14,079. In practice, this meant orders under ₩10,000 were virtually impossible to place at most restaurants. At the time of the announcement, Baemin stated that the implementation schedule and exact subsidy mechanics would be determined later; we were unable to confirm through official notices how these promises were ultimately reflected in merchant settlement statements for this issue.

1 year later, the same headline figure—“₩300 billion over 3 years”—reappeared in Baemin’s consent decree relief package. This time, the package bundled fee reductions for self-delivery merchants, delivery fee subsidies, and a mutual growth cooperation fund. Whether these two pledges overlap cannot be verified from public records. What is documented, however, is the verdict of the Fair Trade Commission (FTC). The antitrust regulator concluded that several proposed measures duplicated existing promotions, lacked operational specificity, or—in cases like subsidies for newly registered stores—bore little connection to the merchants harmed by the platform’s past conduct. The regulator’s stance was unmistakable: regardless of how large the headline pledge looks, where that money actually lands—at which storefronts and against which line items—must be evaluated on its own merits.

Legislative proposals to cap platform fees currently debated in the National Assembly face the exact same dilemma. The 3 bills pending before the National Policy Committee take fundamentally different approaches to defining a ceiling.

Bill	Items Included in Cap	Delivery Fee
Representative Lee Kang-il’s Bill	Sum of brokerage fees, payment processing fees, and ad spend capped within 15% of order revenue	Separate ceiling and floor
Representative Park Jeong-hun’s Bill	Sum of the above plus delivery fees capped within 15%	Included within the 15%
Representative Kim Nam-geun’s Bill	Preferential fee rates for micro- and small businesses instead of a statutory cap	Not applicable

Whether delivery charges are bundled into the calculation transforms the exact same 15% threshold into a completely different operational constraint. The ₩3,400 merchant delivery fee ceiling applied to the top 35% revenue tier alone represents 17% of a ₩20,000 order. Under a bill that includes delivery fees inside the cap, certain orders would breach the legal limit even if brokerage and advertising fees were set to ₩0.

The counterarguments are equally sharp. During the parliamentary audit, lawmakers warned of a “balloon effect,” asking whether capping brokerage commissions would simply drive platforms to raise ad rates or packaging fees; the FTC chairperson testified that if the bill passed, presidential enforcement decrees would plug loopholes to prevent circumvention via ancillary fees. Legal and economic scholars warned that because delivery costs are dictated by distance and transit time rather than ticket size, a percentage-based cap could make it impossible to recoup logistics expenses on small orders. Delivery rider unions voiced concern that capping merchant delivery expenses by law could ultimately depress couriers’ take-home earnings. Merchant associations, conversely, maintain that anything short of an all-inclusive cap covering total platform expenses will leave small business owners unprotected.

Oswarld’s Lens

When a commission cut and a delivery fee hike are announced as a bundle, the commission rate dominates the press release headline, but the delivery fee dictates what the restaurant actually experiences. Platforms tout a “cut of up to 7.8%p,” while restaurant owners insist “nothing has changed”—and both statements can be true at once. They are simply reading different rows of the very same fee schedule.

This is why in the debate over fee caps, definitions matter just as much as a headline figure like 15%. Depending on whether the numerator includes delivery fees and advertising costs, or whether the denominator is calculated before or after promotional discounts, that same 15% can function as a toothless guideline or a rule that effectively bans small-ticket orders. Until the definitions are nailed down, simply arguing over “what percentage” to set will never settle the issue.

Closing

The fee schedule under the win-win plan might look straightforward at first glance, but even the same 7.8% yields entirely different figures depending on which tier you fall into, the total order value, and whether the percentage is applied before or after discounts. A formal ruling on the merits from the Fair Trade Commission (FTC, Korea’s antitrust regulator) and legislative review of statutory fee cap bills are still on the horizon. When those outcomes arrive, it will pay to look first at which line item on your settlement statement actually changes, rather than the headline rate in the press.

---

💬 If you run a business, which item on your recent settlement statements felt heavier than the platform fee itself? Delivery costs, advertising fees, or funding customer coupons—let us know in the comments below.

📨 If you know an entrepreneur or store owner struggling to make sense of delivery app settlements, please share this issue with them.

---

References & Further Reading

Primary sources

• Fair Trade Commission (FTC), “FTC Rejects Consent Decree Application by Baemin and Coupang,” June 18, 2026. (Published on the KDI Economic Information and Education Center) Link ··· Review the original FTC document detailing the decision to reject the consent decree and the full list of investigated charges.
• Bizwatch, “Baedal Minjok Implements ‘Mutual Growth Plan’ in Earnest… Brokerage Fees Cut by up to 7.8%p,” January 22, 2025. Link ··· A clear compilation of tiered fee rates, merchant delivery fees, previous rates, and savings calculations based on an average order value of ₩25,000 (~$18.1). This serves as the baseline for our break-even analysis.
• Herald Economy, “‘Insufficient Restoration of Market Competition’: Baemin and Coupang’s Mutual Growth Plan Rejected… Delivery App Sanctions Loom,” June 17, 2026. Link ··· Detailed reporting on why the FTC rejected the voluntary relief proposal, including relevant revenue figures and the examiner’s estimated penalties.
• Seoul Economic TV, “FTC Rejects Consent Decree for Baemin and Coupang… Sanctions to Be Decided via Formal Deliberation,” June 19, 2026. Link ··· Details the operational timeline of MFN clauses, enforced by Coupang Eats since March 2023 and Baemin since May 2024.
• The Korea Economic Daily, “FTC Initiates Sanctions Against Coupang Eats and Baemin Over ‘Most-Favored-Nation Demands,’” October 13, 2025. Link ··· Lays out the FTC’s rationale identifying MFN clauses as the primary driver behind escalating merchant fees.
• Dailian, “FTC Corrects Unfair Terms and Conditions of ‘Coupang Eats’ and ‘Baedal Minjok,’” October 13, 2025. Link ··· Regulatory findings on unfair clauses, including fees charged on pre-discount prices and inadequate notice regarding merchant exposure radii limits.
• Gyeonggi Shinmun, “SMEs on Delivery Apps Pay 26% of Sales in Platform Costs… Up 5.2%p in 1 Year,” September 2026. Link ··· Coverage of findings from the Korea Federation of SMEs’ (KBIZ) ‘2026 Survey on Transaction Practices of Online Platform Merchants.’
• CEO Score Daily, “The Paradox of Delivery App Fee Caps: ‘Will Saving Merchants Simply Shift the Burden onto Consumers?’” October 14, 2026. Link ··· Analyzes the differences among 3 pending legislative bills, breaks down merchant cost structures, and surveys merchant sentiment regarding tiered fee models.

Background

• MTN, “Delivery Fee Caps Gain Momentum Amid Mixed Hopes and Concerns,” October 13, 2026. Link ··· Covers the National Assembly audit’s famous pork cutlet cost breakdown, along with counterarguments concerning balloon effects, small orders, and rider compensation.
• Kyunghyang Shinmun, “Baemin to ‘Waive Brokerage Fees on Orders Under ₩10,000 (~$7.2),’” June 19, 2025. Link ··· Details the interim agreement supporting small-ticket orders, the share of orders under ₩15,000 (~$10.8), and reactions from merchant associations.
• Maeil Shinmun, Report on Consumer Public Network’s Survey of Delivery App Minimum Order Amounts, June 24, 2025. Link ··· The data source for average minimum order values across delivery platforms.
• Money Today (published on Supple), 2026 National Assembly Audit Report on Coupang Eats’ Fees Charged on Pre-Discount Prices, October 23, 2026. Link ··· Examines platform compliance status 1 year after the corrective recommendation was issued.

---

The author is Oswarld (Kwangseob Ahn). Current roles: Adjunct Professor at Sejong University, Strategy Consultant at INLEVEL9. Career, research, books, and recent work are kept current on the About page. Latest · July 2026: HEMA-2: A Consolidation-Aware Tri-Memory Architecture with Multi-Channel Scheduling for Lifelong Conversational AI.

📝 Glossary

---

Footnotes

1. Baemin Baedal (Baemin1 Plus): A platform-managed delivery model where Baemin handles rider dispatch directly. Unlike “Store Delivery” (gage baedal, where merchants contract independent local delivery agencies directly), merchant delivery fees are deducted straight from Baemin’s settlement payouts. Tiered fee rates under the mutual growth plan apply specifically to this model. ↩

2. Most-Favored-Nation (MFN) Demand: A contractual clause requiring a counterparty not to offer less favorable terms than those offered to competitors. In delivery apps, this manifests as barring merchants from setting higher menu prices, higher minimum order amounts, or smaller discounts than what they offer on competing platforms. ↩

3. Consent Decree (dong-ui uigyeol): A legal mechanism where a business facing antitrust charges proposes voluntary remedies and consumer relief measures. If the FTC approves them, the case is resolved without formally ruling on the illegality of the conduct. If rejected, the case returns to standard formal adjudication. ↩

4. Relevant Revenue: The total sales revenue generated from business activities directly tied to the alleged legal violation. Administrative fines are calculated by applying a penalty rate to this figure based on the gravity of the infraction; abuse of market dominance carries fines of up to 6% of relevant revenue. ↩