What actually changed about Korean delivery-platform fees in 2026
A year and a half after the Korea Fair Trade Commission's (KFTC) 2024 recommendation, both Baemin and Coupang Eats now charge 9.8% brokerage. Restaurant owners still feel the burden: the 9.8% is up from the prior 6.8%. Yogiyo sits at 12.5% on paper — the highest — but its mix of owner-delivery options can make the real take-home closer than the headline suggests.
This guide compares the three platforms using KFTC 2025 online-platform fee disclosures, each platform's official seller-center material, and operating patterns from restaurants averaging around KRW 25M in monthly revenue. The conclusion is simple: delivery-cost share and ad efficiency move owner profit more than the headline brokerage rate.
2026 fees at a glance
| Item | Baemin (Baemin1+) | Coupang Eats | Yogiyo |
|---|---|---|---|
| Brokerage fee | 9.8% | 9.8% | 12.5% |
| Payment fee | 3.0% | 3.0% | 3.5% |
| Delivery share (owner) | KRW 1,900–2,900 | KRW 2,900–3,900 | KRW 0–3,500 |
| Ads (e.g., Baemin "flag") | KRW 88K/mo (base) | CPC bid | CPC bid |
| Payout cycle | D+1 | D+1 | D+3 |
| VAT handling | Separately billed | Separately billed | Separately billed |
| Rider pool | Own + partner | Own (Coupang Eats riders) | Partner-heavy |
The 2024 KFTC recommendation asked operators to refrain from passing fee increases through to consumer pricing, but field data from the Korea Foodservice Industry Association and the Korea Self-Employed Coalition (Nov 2025) suggests owner net margin still dropped about 2–3%.
KRW 15,000 menu simulation
Using KRW 15,000 as a typical ticket size, with COGS at 35% and average delivery share, the three platforms produce different net profits.
| Line | Baemin | Coupang Eats | Yogiyo |
|---|---|---|---|
| Revenue | 15,000 | 15,000 | 15,000 |
| Brokerage | −1,470 | −1,470 | −1,875 |
| Payment | −450 | −450 | −525 |
| Delivery share | −2,400 | −3,400 | −2,000 |
| COGS (35%) | −5,250 | −5,250 | −5,250 |
| Net | 5,430 | 4,430 | 5,350 |
Yogiyo's higher headline fee is offset by its KRW 0 owner-delivery option, so the real spread between platforms narrows. Coupang Eats carries the heaviest delivery share, so low-ticket menus deteriorate quickly there.
Use the Baemin Profit Calculator and Delivery Platform Comparator to plug in your menu mix and average ticket — the platform-priority answer becomes immediate.
Ad structures — apples to oranges
Beyond the fee table, ad cost moves the bottom line, and the ad models are structured so differently that direct comparison is unintuitive.
Baemin
- OpenList — free exposure in search results
- UltraCall — KRW 88K/mo per region, exclusive top-slot is not available
- Baemin1+ Flag — extra exposure with the same fee, additional flag-operation cost separate
Coupang Eats
- CPC bid — average KRW 700–1,500 per click, large variance by category and time slot
- Coupang Eats Boost — fixed fee combined with higher take rate
- New-seller 50% ad credit — first 3 months, ends once revenue passes a threshold
Yogiyo
- CPC bid + flat ads — both available concurrently
- KRW 0 delivery policy — when the owner sets KRW 0, Yogiyo absorbs the delivery cost but exposure priority is slightly lower
- Yogi-Pass — subscribers see priority listings, no additional cost to the owner
Costs owners often forget
Beyond fees and delivery share, several costs compound silently. Build them into your operating P&L from day one.
- Menu photography — KRW 100K–500K per session, repeat shots needed for menu refreshes
- Packaging and labels — KRW 200–500 per item, hard to pass on as price hikes
- Review events — free side dishes or coupons, on average 5–10% of menu price
- VAT — 10% of settled revenue, separate from platform statement (general taxpayers)
- Cash receipt / tax invoice handling — auto on some platforms, manual on others
- Rider tips — unofficial, but some owners pay for priority dispatch
Matching restaurant type to platform
Concept, ticket size, and operating hours all shift the platform recommendation. Single-owner operating load matters too.
| Restaurant type | First pick | Reason |
|---|---|---|
| KRW 15K+ Korean / Chinese | Baemin + Coupang Eats | Ticket supports both, dual exposure |
| Café / dessert | Baemin only | Low ticket struggles with Coupang Eats BEP |
| Late-night / chicken / pizza | All three (time-segmented) | Exploit time-segmented rider supply |
| Solo kitchen | Baemin + Yogiyo | KRW 0 delivery simplifies ops |
| Snack box / lunch box | Baemin first | Lunch-hour ad efficiency is strongest |
| Japanese / sushi | Coupang Eats + Baemin | KRW 20K+ ticket favors Coupang Eats |
| Bakery | Baemin only or Yogiyo | Fresh-food category exposure |
Time-of-day playbook
Traffic differs significantly across day parts. Typical operating patterns look like this.
- Lunch (11:30–14:00) — Baemin dominates, office-district lunch-only menus expose well
- Dinner (17:30–20:30) — Coupang Eats leverages rider-pool advantage, attracts fast-delivery customers
- Late night (21:00–02:00) — Yogiyo / Baemin share parity, chicken / pizza / late-night categories peak
- Weekend lunch / dinner — all three peak; ad CPCs run 1.3–1.5× weekday levels
Payout cycle and working capital
All three platforms settle within 1–3 business days of revenue recognition. Restaurants buy ingredients daily, so faster settlement materially helps working capital.
| Platform | Cycle | Weekend / holiday handling |
|---|---|---|
| Baemin | D+1 | Weekend revenue settles Monday |
| Coupang Eats | D+1 | Same |
| Yogiyo | D+3 | Business-day basis |
The headline gap looks small, but for a KRW 30M/month restaurant, D+3 means an average of KRW 300K–500K perpetually locked in transit.
Review and rating management — drives ~30% of revenue
All three platforms boost top-of-search exposure significantly when ratings exceed 4.5. Review management is therefore the most direct way to reduce ad spend.
- First 5 reviews — a new restaurant with zero reviews effectively converts at near-zero. Pair "free side dish + review request" through the first 3 weeks to hit 5 reviews as a first milestone.
- Negative review response — 1–2 star reviews need a response within 24 hours. The standard template is apology + recurrence-prevention commitment + invitation to return.
- Photo review incentive — photo reviews carry ~1.5× exposure weight. "KRW 1,000 off next order for a photo review" is an explicit and effective incentive.
- Malicious review takedown — clearly defamatory reviews can be reported via each platform's flag system. Baemin has the highest takedown processing rate.
Government and platform support programs
Since 2025, several programs have been running to ease small-merchant burden. Many require explicit application by the owner — easy to miss.
- Small-merchant payment-fee refund — businesses under KRW 3B annual revenue receive partial refunds on payment fees. Apply via the Small Enterprise and Market Service portal.
- Baemin / Coupang Eats ad discount — 50% ad-credit for first 3 months, occasionally raised to 60% during region-activation campaigns.
- Yogiyo KRW 0 delivery subsidy — partial revenue guarantee when running the KRW 0 owner-share option for a defined period.
- Self-employed unemployment insurance — voluntary enrollment, provides safety net against revenue volatility.
FAQ
- Q. Should I onboard all three? → If your average ticket is KRW 15K+ and menu turns over fast, three-platform onboarding pays off. Solo owners often start with two to cap CS and inventory load.
- Q. Does the KRW 0 owner-delivery option really lift volume? → Average order count rises 1.2–1.4× in field data. Low-ticket cafes / desserts struggle to absorb it though, so simulate before adopting.
- Q. Can I run with zero ad spend? → New restaurants effectively must run ads for the first 3 months. After 6 months, if regulars account for 30%+ of orders, ad spend can taper.
- Q. Own delivery vs platform delivery? → Past about KRW 50M/month, hiring one in-house rider becomes feasible. Below that, platform riders are more efficient.
- Q. How do I handle VAT? → General taxpayers file in January and July. The platform settlement statement is usable as revenue evidence. Keep all ingredient-purchase receipts for input VAT credit.
Related tools
- Baemin Profit Calculator — combines brokerage, payment, delivery share, and COGS
- Delivery Platform Comparator — simulate the same menu across all three
- SmartStore Fee Calc — comparison reference for non-delivery sales
Bottom line
With 2026 fees converged, the levers that matter are owner delivery share, ad ROI, and time-of-day operations. A KRW 1,000 per-order net-profit gap compounds to about KRW 1M/month and KRW 12M/year.
The safest playbook is to run different platforms by menu and time slot. Lunch menus on Baemin, dinner menus on Coupang Eats, late-night menus on Yogiyo. Reserve three-platform exposure for tickets above KRW 15K — and cap at two platforms for lower-ticket menus to keep operating load sane.