Why GMV and ROAS are not profit
GMV and ROAS measure sales, not profit: neither subtracts what it costs to make, sell and deliver those sales.
GMV adds up the value of orders. It does not take off TikTok's fees, creator commission, product cost or shipping. ROAS divides sales by ad spend. Our research on paid creator video describes estimated ROAS for a TikTok Shop video as attributed revenue before product cost, fees, shipping, returns and other expenses, “so it is not profit.”
Our own case studies report two measures: TikTok Shop GMV and blended ROAS. On those pages, blended ROAS is total TikTok Shop GMV for the program divided by the TikTok ad spend managed over the program window. The Terra Health case study, for example, reports $30K in managed TikTok ad spend and $323.2K in TikTok Shop GMV, a blended 10.8x ROAS.
Those numbers show how much a program sold and how large that was against the paid budget. They do not show profit, because product cost, fees, commission, samples, shipping and refunds are left out. They do not show incrementality, because the GMV includes organic and creator sales that might have happened anyway. Read them as evidence of sales scale, not of margin.
Contribution profit, line by line
Contribution profit is GMV minus every cost that rises and falls with sales: refunds, TikTok fees, creator commission, samples, fulfilment, product cost and ad spend.
The TikTok Shop Forecaster works this way. Its monthly P&L runs from GMV through returns, product cost, shipping and fulfilment, creator samples, TikTok fees and creator commission to contribution, then subtracts the ad budget to give profit after ads. We recommend deducting any agency fee as a separate line after that, before calling the result profit.
| Line item | What it is | Where the number comes from |
|---|---|---|
| GMV | The value of orders placed, at the price customers paid | Seller Center order reports, at the price after discounts, not the list price. |
| Discounts and refunds | Discounts lower the price paid; refunds reverse revenue and can add shipping or handling costs | Settled-order reports and your return policy. Our fees guide lists 2026 benchmark return rates from 4% in food and beverage to 25% in footwear. |
| TikTok referral fee | The platform fee on qualified orders, which also covers payment processing | Usually 6% in the US, with category exceptions. Check the category schedule in Seller Center. |
| Smart Promotion fee | An enrollment-based fee on store GMV | 3.5% outside campaigns, with campaign increases, while you are enrolled. |
| Creator commission | The rate you set per product and collaboration, paid on affiliate orders | Your affiliate settlement report. Canceled or fully refunded orders earn no commission. |
| Samples | Free product, and the shipping on it, sent to creators | Your sample log: units sent multiplied by product cost plus shipping. |
| Fulfilment and shipping | Pick, pack and ship per unit, plus storage | Your FBT or 3PL invoices. The forecaster assumes Fulfilled by TikTok typically costs $4.14–$4.80 a unit, all in. |
| Cost of goods | The landed cost of each unit sold | Your own accounts. |
| Ad spend | GMV Max, Spark Ads and other TikTok ad spend in the same period | TikTok ads invoices for the period. |
| Agency fee | Management fees | Your agency contract. |
TikTok calculates the referral fee on customer payment plus platform discount minus tax, and Smart Promotion on store GMV; the forecaster applies both to net sales. Sales tax on fees is excluded.
Worked example: one month on TikTok Shop
Illustrative example, not a client result. It uses the same $32 order as the margin table in our TikTok Shop fees guide, with that table's inputs: 6% referral and 3.5% Smart Promotion on all GMV, creator commission at a 20% rate on the 70% of sales that are affiliate-attributed, $3 fulfilment a unit, a 6% revenue-loss allowance for returns, $8 product cost and ad spend equal to one-third of GMV. As in that table, fees and commission apply to GMV and returns are counted as lost revenue. The rest are example inputs: 10,000 orders and 200 samples.
| Line | Calculation | Amount |
|---|---|---|
| GMV | 10,000 orders × $32 | $320,000.00 |
| Referral fee | 6% × $320,000 | −$19,200.00 |
| Smart Promotion fee | 3.5% × $320,000 | −$11,200.00 |
| Creator commission | 20% rate × 70% affiliate share × $320,000 | −$44,800.00 |
| Fulfilment | 10,000 × $3 | −$30,000.00 |
| Returns | 6% × $320,000 | −$19,200.00 |
| Samples | 200 × ($8 + $3), example input | −$2,200.00 |
| Cost of goods | 10,000 × $8 | −$80,000.00 |
| Contribution before ads | 35.4% of GMV | $113,400.00 |
| Ad spend | $320,000 ÷ 3; blended ROAS 3.00x | −$106,666.67 |
| Profit after ads | 2.1% of GMV | $6,733.33 |
| Blended ROAS | Ad spend | Profit after ads | Share of GMV |
|---|---|---|---|
| 3.00x | −$106,666.67 | $6,733.33 | 2.1% |
| 4.00x | −$80,000.00 | $33,400.00 | 10.4% |
| 5.00x | −$64,000.00 | $49,400.00 | 15.4% |
| 6.00x | −$53,333.33 | $60,066.67 | 18.8% |
| 8.00x | −$40,000.00 | $73,400.00 | 22.9% |
At the fees guide's 3.00x blended ROAS this shop keeps $6,733.33, or 2.1% of GMV, after ads; at 5.00x it keeps $49,400.00, or 15.4%. Break-even is $320,000 ÷ $113,400 = 2.82x. To include an agency fee, subtract it from contribution before ads first: break-even ROAS = GMV ÷ (contribution before ads − agency fee). Work out your own break-even ROAS before you judge any ROAS number. Fulfilment of $4.50 a unit, inside the forecaster's Fulfilled by TikTok range, would be about 14% of a $32 order and about 25% of an $18.10 order, so lower-priced products need bundles or cheaper shipping to reach the same margin.
Attributed or incremental
Attributed sales are the orders TikTok links to a creator, video, LIVE or ad; incremental sales are the orders that would not have happened without them, and the two can differ a lot.
TikTok attributes an affiliate order to the creator when a viewer buys through that creator's link or tagged content. Ads reporting, in the words of our research, “assigns sales to ads based on who saw and then bought.” Neither rule asks whether the buyer would have bought anyway.
Some attributed orders would have happened without TikTok:
- Existing customers who already knew the brand and bought on TikTok because the product was in their feed.
- Moved orders that would otherwise have gone through Amazon or the brand's own store.
- Organic demand in paid reporting. Promotion tends to follow videos that already sell, which, as our paid UGC measurement paper puts it, would inflate attributed returns “even if the ads added little.”
How to test it
We recommend these checks:
- New-customer share. Where Seller Center or your own customer data separates first-time from repeat buyers, report the share of TikTok Shop orders from new customers each month.
- Holdouts for paid. Hold back paid promotion in randomly chosen regions, audiences, products or time blocks, keep the rest running, and compare total sales rather than attributed sales. Set the decision rule, on contribution margin, before you start. Our agency audit research sets out a holdout by products or time blocks.
- Amazon and own-store before and after. Compare Amazon branded search and sales, and your own store's sales, before, during and after a TikTok push, against the same weeks a year earlier. This is weaker than a holdout, because other things change too.
- Paid and organic apart. Report organic and paid views and sales on separate lines for every video, as our agency reporting research recommends.
- Incremental ROAS and incremental contribution. Divide incremental revenue by incremental ad spend, then estimate incremental contribution after product cost, fees and returns, and deduct the incremental ad spend.
A GMV Max target ROI is a bar you set for the campaign, not a measured result; our TikTok ads agency page covers GMV Max, Spark Ads and LIVE promotion.
Halo: sales that show up on Amazon and in retail
Halo is extra revenue that TikTok activity drives on channels TikTok Shop does not record, such as Amazon and retail stores; the forecaster estimates it from third-party test results, and only your own data can confirm it for your brand.
The forecaster's assumption, as published: “Haus’s geo-holdout tests found $0.34 more on Amazon and $0.57 more in retail stores for every $1 TikTok drives online. We show half to the full figure, on the channels you pick. It isn’t in GMV or profit.” On $100,000 of TikTok online sales, that range is $17,000 to $34,000 on Amazon and $28,500 to $57,000 in retail stores.
Treat that range as a planning estimate: the ratios come from Haus's tests, not your brand. Halo is also revenue, not profit; to count it in profit, apply that channel's own margin, fees and costs.
The forecaster also shows earned media value: creator video views, about 100 for every $1 of GMV on the accounts we run, priced at an $18 CPM. It is a media value, not revenue, and is not in profit.
To check halo against your own numbers:
- Compare Amazon branded search and units in weeks with heavy TikTok activity against weeks without it, and against the same weeks a year earlier.
- Where you have retail sales by region, compare regions with more TikTok activity against regions with less, or use a paid holdout by region.
- Divide the extra sales by your TikTok Shop sales over the same weeks, and use that ratio instead of the forecaster's.
What a monthly TikTok Shop report should show
We recommend that a monthly TikTok Shop report show sales by channel, every cost line down to contribution profit, and the tests behind any claim of incremental or halo sales.
- GMV by channel
Shoppable video, LIVE, the shop tab and ads, on separate lines. Our sales channels research shows how that split varies by category.
- Paid and organic
Views and sales for each, by video and in total.
- Refunds and returns
Value and rate, by product.
- Fees, commission and samples
TikTok referral and Smart Promotion fees, creator commission by type, and sample units and cost.
- Ad spend
Actual spend, blended ROAS and the break-even ROAS implied by your margin.
- Contribution profit
Before and after the agency fee, calculated only once every cost line is present.
- New customers
The share of orders from first-time buyers, where the data exists.
- Halo indicators
Amazon branded search and sales, own-store sales and retail sell-through where available.
- Tests run and results
What was held out, for how long, the result and how uncertain it is.
Our mid-market programs include halo measurement, and our enterprise programs include cross-channel halo modeling across D2C, Amazon and retail. Our social commerce agency page explains why we focus on TikTok Shop.
What this method cannot tell you
This method gives a defensible estimate of TikTok Shop profit, but it has limits that no report should hide.
- Attribution is not cause. Without a holdout, the incremental share of sales is a judgment, not a measurement.
- Halo ratios are borrowed. The forecaster's halo comes from Haus's tests, not your customers.
- One month is not a lifetime. Repeat purchases may improve lifetime economics, but they should not be assumed to rescue an unprofitable first order. Measure contribution by customer cohort.
- Fixed costs and tax sit outside. Contribution profit excludes staff, software, overheads and sales tax on fees.
- Inputs change. TikTok adjusts fees and programs, and third-party market data is estimated, not reported by TikTok. Check Seller Center before relying on a rate.
FAQ
Put your price, product cost and commission rates into the free TikTok Shop Forecaster to see a six-month GMV forecast in three scenarios with a halo range and earned media value, then unlock profit after ads and the monthly P&L.
Model your own numbersAuthor: Lukas Minkov, Co-Founder & CEO at MediaLabs. LinkedIn.
Last updated: November 10, 2026.