Profit
TikTok Shop's commission structure is the only one of the three SEA marketplaces with a stacked second commission layered on top of the platform fee. TikTok Shop's Seller University documents Affiliate Plan commissions in the 1–80% range (seller-set per SKU), with the Open Affiliate Plan defaulting to 10–25% on most categories. Add the platform commission (typically 1–5% in Thailand depending on category), the seller-funded voucher tiers required to maintain ranking in live-commerce surfaces (8–12% during peak campaign windows), the platform shipping subsidy, and a returns reserve elevated above Shopee or Lazada by the impulse-purchase profile of the buyer base, and the structural cost-per-order is materially higher than the platform-reported ROAS suggests.
TikTok Shop's commission stack has four lines, not one. Platform-reported ROAS counts only the first.
This post argues that the popular TikTok Shop advice — run more live-streams, enroll in the Open Affiliate Plan, follow what the recommender rewards — is structurally biased toward overspend. Live-stream GMV is not margin. Affiliate-driven sales arrive with a 10–25% commission stacked on top of the platform fee. The fix is not to retreat from these surfaces; it is to set per-SKU affiliate commissions, price live-stream SKUs against margin instead of session lift, and reconcile contribution margin from order-line data with the full four-line stack subtracted.
The four-line cost stack — visualised
The chart below decomposes a representative THB 30,000 ad-attributed TikTok Shop sale through the typical Thai LazMall-equivalent SKU profile. The brand-orange bar at the bottom is the contribution margin the seller actually has. The arithmetic gap between the Shopee-style cost stack (one commission line) and the TikTok Shop stack (commission plus affiliate plus elevated returns) is the entire reason TikTok Shop margin economics behave differently from the other two platforms.
Two highlighted lines distinguish TikTok Shop from Shopee and Lazada economics: the affiliate commission (THB 6,000 here, ~20% of revenue on the Open Plan default) and the elevated returns reserve (THB 1,800, ~6% of revenue vs. ~2–3% baseline on Shopee/Lazada). Together they account for ~26% of revenue — the structural cost gap that turns a 6.0 platform ROAS into a 1.68 true ROAS.
A worked example: the 6.0 platform ROAS that is barely break-even
Platform view:
Ad-attributed revenue: THB 30,000
Ad spend: THB 5,000
Platform ROAS: 6.0 ← dashboard says winner
True ROAS view (affiliate-tagged orders, typical category):
COGS (35%): THB 10,500
Platform commission (3%): THB 900
Affiliate commission (20%): THB 6,000
Seller-funded voucher (5%): THB 1,500
Free-shipping subsidy (3%): THB 900
Returns reserve (6%): THB 1,800
Ad spend: THB 5,000
Contribution profit before ads = 30,000 − 21,600 = THB 8,400
Contribution profit after ads = THB 3,400
True ROAS = 8,400 / 5,000 = 1.68The campaign is profitable but only barely. The ratio collapsed from 6.0 to 1.68 once the full cost stack was applied. A 5-percentage-point increase in returns reserve (typical for impulse-prone live-stream SKUs) flips it under 1.5. A 5-percentage-point increase in affiliate commission (a common Open Affiliate Plan default for higher-tier creators) does the same. A seller optimising on platform ROAS has no signal that anything has changed.
Affiliate commission is the line most accounts under-manage
TikTok Shop's Open Affiliate Plan is opt-in but default-enrolled for most new sellers. Once enrolled, creators discover the SKU and earn the published commission rate on every attributed order. On accounts that have not actively configured per-SKU rates, 40–60% of order volume routinely flows through the Open Plan at a 15–25% commission — stacked on top of the 1–5% platform commission. For high-margin SKUs (50%+ contribution margin) this is sustainable; for thin-margin SKUs it is the difference between profit and loss on every affiliate-tagged order.
The fix is per-SKU affiliate commission tiering: set the Open Plan commission per SKU based on the SKU's actual margin, gate high-margin SKUs (the ones that fund the catalog) out of the Open Plan and into the Targeted Plan with curated creator partners, and review the mix monthly. Most accounts that do this once recover 3–5 percentage points of margin without losing meaningful affiliate-driven volume — the creators who matter for the SKU stay engaged through the Targeted Plan; the long tail of opportunistic affiliates is throttled.
Live-stream pricing: the host's incentive is not the seller's
Live-stream hosts price for the spike. They ask for aggressive discounts because aggressive discounts drive in-session conversions, and the host's affiliate commission scales with the GMV produced during the session. The seller's incentive is the margin that survives the session. Both incentives are real; they are also structurally misaligned. The seller cannot delegate live-stream pricing to the host without losing margin discipline.
Two operating rules. First, set a discount floor per SKU — the maximum discount the SKU's margin can absorb at break-even, with a small buffer. Hosts work within the floor. Second, rotate which SKUs are featured: the highest-discount slot in a session should rotate across SKUs, not always land on the highest-margin product. This keeps the host's spike-driving discount available without burning the same margin lines repeatedly. Sellers who run this discipline find that 3–5 live-stream sessions per month account for the bulk of their margin compression, and that two changes (tier the discount, rotate the featured SKU) recover most of it.
Same SKU, same campaign, same dashboard ROAS. As the Open Affiliate Plan commission rises from 5% (curated low-tier creators) to 30% (high-tier creator marketplace), true ROAS declines roughly linearly — and crosses break-even (~1.0) somewhere between 22% and 28% commission depending on category margin. The chart is the operational case for setting affiliate commission per SKU: the same affiliate program is profitable on a high-margin SKU and loss-making on a thin-margin one at the same commission tier.
Sensitivity — what changes the conclusion
The four-line cost stack varies meaningfully by category, by Affiliate Plan configuration, and by campaign-window timing. The table below stress-tests the worked example.
| Adjustment | New true ROAS | Conclusion |
|---|---|---|
| Baseline (worked example) | 1.68 | Profitable, thin buffer |
| Affiliate commission 20% → 25% (Open Plan default drift) | 1.32 | Approaches break-even — common Open Plan default for higher-tier creators |
| Returns reserve 6% → 9% (live-stream impulse profile) | 1.50 | Returns reserve is the most under-modelled line on most accounts |
| Voucher 5% → 10% (peak-campaign tier) | 1.38 | Campaign-window stress on top of affiliate stack |
| Combined: affiliate 25% + voucher 10% (peak Open Plan campaign) | 1.04 | Break-even — typical 11.11 / 12.12 stress profile |
| Combined: affiliate 30% + returns 9% + voucher 10% | 0.62 | Severe loss — high-tier creator on impulse-prone SKU at peak |
| Move SKU to Targeted Plan at 12% (curated creators) | 2.05 | The recovery move — gates the SKU out of Open Plan defaults |
The table makes the structural argument: the Affiliate Plan commission tier is the highest-leverage seller-controlled lever in the TikTok Shop cost stack. Combined with returns-reserve discipline and voucher-tier discipline, per-SKU affiliate tiering is the difference between 1.0 and 2.0 true ROAS on the same SKU.
TikTok Shop's economics are not Shopee's with a different platform skin. The affiliate stack and the returns profile change the math. Reuse the framework, recalibrate the inputs.
Limitations and where this argument breaks
- Account-size lower bound. The framework's recovery scales with affiliate-channel exposure. Smaller TikTok Shop accounts with low affiliate-tagged volume see less marginal benefit from per-SKU Affiliate Plan tiering; below ~THB 200K monthly revenue, simpler heuristics (uniform Open Plan with conservative voucher tiers) outperform the operational overhead.
- Creator-marketplace dynamics. The 10–25% Open Plan default range is documented in TikTok Shop Seller University, but actual creator participation varies by category and account profile. Categories with low creator interest (industrial, B2B) see less affiliate traffic; categories with strong creator participation (beauty, apparel, home) see commission tiers concentrated near the upper end of the range.
- Live-stream session attribution. Margin reports run at the campaign-aggregate horizon understate the contribution of individual live-stream sessions, because session-attributed orders mix click-time with order-close attribution. Reconcile session-level economics on a 14-day post-session window for accuracy.
- Returns-reserve elevation. The 4–8 percentage-point returns elevation above Shopee/Lazada baseline is observed in our sample but varies by category. Apparel and beauty (impulse-prone) skew higher; consumables and homeware (utility-driven) skew lower. Recalibrate the reserve per category before applying the framework as a bright-line.
- Time horizon. TikTok Shop is the youngest of the three SEA marketplaces and its commission structure has shifted multiple times since launch. The framework is robust; the specific commission tier and Open Plan defaults need quarterly refresh against Seller University documentation.
- Internal-data scope. The 40–60% Open Plan share, the 3–5 sessions/month margin-compression concentration, the typical Affiliate Plan commission distribution — all aggregated across the SEA-6 Thai TikTok Shop accounts we model directly. They are not population claims about all TikTok Shop sellers; they exclude operators below the size bound and the cross-border seller cohort.
Methodology
Public-data citations are taken from the TikTok Shop Seller University documentation (Affiliate Plan commission, platform commission and transaction fee, live-stream session and creator marketplace docs), ByteDance newsroom announcements on TikTok Shop SEA expansion, the Bain e-Conomy SEA 2025 commentary on video-commerce growth, and the Google–Temasek–Bain e-Conomy SEA 2025 macro report.
Internal-data claims — the 40–60% Open Plan share, the 6.0 → 1.68 true-ROAS collapse on a typical category, the 3–5 sessions/month margin-compression pattern, the typical Affiliate Plan commission distribution — are aggregated across the Thai SEA-6 TikTok Shop accounts we model directly. The TikTok Shop subset comprises approximately 120 active accounts across the DataGlass research methodology sample frame (Jan 2024 – Apr 2026, 28-month observation window).