Influencer marketing ROI on TikTok Shop: what the data shows and how to measure it

Updated October 21, 2026 · ~12 min read · By the MediaLabs team

Below: why the numbers exist on TikTok Shop, the formula, a worked example with hypothetical figures, what MediaLabs Research data (in collaboration with Coherence Limited) shows about creator-led sales, and the results our own clients have published. Reported ROAS and real return are not the same thing, and the gap gets its own section.

About the data: research figures are estimates for top-selling US videos and products (tracked best-sellers, about 100 per category). They are observational, not TikTok's official figures, and they describe winners, not typical videos. Each table states its window.

Why creator ROI is measurable on TikTok Shop

Most influencer marketing is measured through proxies: reach, engagement, discount codes, brand-lift surveys. The purchase happens somewhere else, often days later. On TikTok Shop the viewer buys inside the app, from the video, LIVE or shop page they are already on.

Each affiliate sale is recorded as an order with a commission attached. TikTok's affiliate orders FAQ gives the calculation, (revenue − refunds) × commission rate, and commission is paid after delivery. The order-level data you need for ROI already sits in Seller Center.

Source: winning video economics, 16,541 videos tracked in 29 categories, 3 Sept–2 Oct 2026. MediaLabs Research, in collaboration with Coherence Limited.

Creators carry most of it. In a later window (20 Sept–19 Oct 2026), affiliate creators were credited with 74% of latest-week revenue across categories with channel data, brands' own accounts with 11.4% and the rest with 14.6% (affiliate commission guide).

What this doesn't capture: sales a TikTok video drives on Amazon, your own site or in stores, and orders that would have happened without the creator or the ad. Both need separate measurement, covered further down.

How to calculate influencer marketing ROI on TikTok Shop

Start with the inputs. The last four are paid whether or not sales follow, which makes them the easiest to leave out.

Then five formulas:

• Contribution = GMV − refunds − product cost − platform fees − fulfillment − commission
• Program spend = samples + creator fees + ad spend + team or agency cost
• ROI = (contribution − program spend) ÷ program spend
• Blended ROAS = GMV ÷ ad spend
• Break-even blended ROAS = GMV ÷ (contribution − program spend other than ads)

ROAS tells you how much GMV came in per ad dollar. ROI tells you whether the program made money after every cost. A program can show a high ROAS and still lose money, which is why the example below reports both.

A worked example (hypothetical numbers)

Round numbers chosen for illustration. Not a benchmark, a forecast or a MediaLabs client result.

Program spend is $33,250 ($2,250 + $5,000 + $20,000 + $6,000), so:

• ROI = $12,350 ÷ $33,250 = 37.1%
• Blended ROAS = $100,000 ÷ $20,000 = 5.0x
• Break-even blended ROAS = $100,000 ÷ ($45,600 − $13,250) = about 3.09x

The same month reads as a 5.0x ROAS and a 37% return. At this cost structure, a blended ROAS under about 3.1x would lose money before overheads. And even the 37% assumes every attributed order was caused by the program, which the incrementality section below questions.

What the data says about TikTok Shop creator ROI

These are the figures most relevant to creator ROI from MediaLabs Research, in collaboration with Coherence Limited. All cover US TikTok Shop and top-selling videos or products only.

A video-week is one video's sales in one week it ranked among the top sellers. Paid shares are reported per video and capped at 80%, so they are lower bounds. ROAS is estimated revenue per ad dollar, not profit or incremental revenue.

Three things in this table matter for ROI:

• Measure long enough. More than a quarter of top-video revenue came from videos over 30 days old, and older videos earned more per view. A one-week ROI read understates what a creator video returns.
• Judge the portfolio, not single posts. The top 10% of video-weeks took between 24.6% and 44% of category revenue. A few videos carry a lot of the return.
• Paid support goes with scale, not better conversion. Organic and paid-led winners earned similar revenue per 1,000 views; paid-led ones sold about twice as much per week.

Yield and paid promotion by category

Revenue per view varies more than threefold between categories, so benchmark against your own category, not the market.

Sources: winning video economics (revenue per 1,000 views) and paid vs organic (paid share, ROAS). Paid shares are lower bounds. ROAS is attributed, not incremental. MediaLabs Research, in collaboration with Coherence Limited.

Paid vs organic: what reported ROAS does and doesn't tell you

Reported ROAS is attribution, not proof of cause. TikTok's own Product GMV Max help page says all orders from products you advertise are attributed to the campaign, including those from organic content and affiliate orders, and that GMV for same-day orders is calculated with a one-day attribution window. A GMV Max ROI figure therefore includes sales your creators might have made without the ads.

Our newer paid UGC measurement paper (20 Sept–19 Oct 2026, 53,532 videos tracked) shows the same shape at larger scale. Paid-led videos, with 80% or more of revenue paid, took 66.2% of top-video revenue on 57% of seven-day video snapshots. Organic videos took 11.8% on 21.1%. Median ROAS sat between 2.33 and 2.42 in every paid band, and organic videos earned slightly more per 1,000 views than paid-led ones ($17.56 vs $16.17).

Measuring the true return is hard even with experiments. In 25 large field experiments with US retailers and brokerages, Lewis and Rao (Quarterly Journal of Economics, 2015) found the median confidence interval on ROI was more than 100 percentage points wide. Observational ROAS, the kind every dashboard shows, is weaker evidence still.

The fix is a holdout test. The paper proposes one; it is a recommendation, not a finding:

1. Pick a product with stable demand and enough sales to detect a change.
2. Set a margin-based decision rule before you start.
3. Randomly split comparable audiences or regions into a paid group and a holdout group.
4. Keep price, creator posting, discounts and other channels as steady as you can, for a full purchase cycle, and don't stop early on a good-looking result.
5. Compare total sales per group, including organic and other-channel sales. Incremental ROAS = incremental revenue ÷ incremental ad spend.
6. Report uncertainty ranges, and repeat on a second product before scaling.

For running the campaigns themselves, see our GMV Max guide.

Published MediaLabs case studies

MediaLabs is a TikTok Shop agency, and these are programs we ran. Each figure is one brand's result exactly as published on our case-study pages. They are individual case studies, not averages or typical outcomes. Our case studies report blended ROAS: total TikTok Shop GMV for the program divided by the TikTok ad spend managed over the program window, for the TikTok Shop channel only.

Read them with the worked example in mind. Blended ROAS divides GMV by ad spend only. It doesn't subtract product cost, fees, commission, samples or creator fees, and it doesn't separate incremental sales. The spread, from 3.39x to 36.1x, is the more useful lesson: results depend on category, product, price and creator fit. Full write-ups are on our case studies page.

How to measure your own creator ROI

1. Fix the cohort and window. Which products, which creators, which dates. Keep counting for at least 30 days after the last post.
2. Pull attributed GMV and refunds from Seller Center, using settled orders where you can.
3. Book every cost, including samples that never turned into posts, flat fees, ad spend and team time.
4. Report three numbers side by side: blended ROAS, contribution and ROI.
5. Compare revenue per 1,000 views with your category to see whether content or reach is the constraint.
6. Run a holdout before scaling ad spend, so you know the incremental return, not just the attributed one.

To model the numbers before you spend, the free TikTok Shop Forecaster projects six months of GMV, P&L and break-even for your category, with managed creators, open collaboration and GMV Max as inputs.

Frequently asked questions

What is a good ROI for influencer marketing on TikTok Shop?

It depends on your margin. A useful floor is your break-even blended ROAS: GMV divided by what's left of contribution after non-ad costs, about 3.1x in our hypothetical example. For context, paid top-selling TikTok Shop videos reported a median ROAS of 2.42 from 3 September to 2 October 2026, and 25.4% reported under 2.

How do you measure influencer marketing ROI on TikTok Shop?

Take attributed GMV from Seller Center, subtract refunds, product cost, platform fees, fulfillment and commission to get contribution, then subtract program spend (samples, creator fees, ad spend and team or agency cost) and divide the result by that spend. Measure for at least 30 days after the last post.

What is the difference between ROAS and ROI?

ROAS is GMV divided by ad spend; it shows how much revenue came in per ad dollar. ROI subtracts every cost, including product cost, fees, commission, samples and creator fees, and shows whether the program made money. A program can have a high ROAS and a negative ROI.

Is influencer marketing on TikTok Shop worth it?

It is where most TikTok Shop sales happen: shoppable video carried 64.9% of revenue in the latest week to 29 September 2026, and affiliate creators were credited with 74% of latest-week revenue in the 20 September to 19 October 2026 window. Whether it pays for one brand depends on margin, category and creator fit, so measure contribution, not views.

How long does a TikTok Shop creator video keep selling?

Most revenue comes early: the revenue-weighted median age of a selling top video was 11 days. But videos over 30 days old still brought in 27.6% of top-video revenue, and videos over 90 days old 13.3% (3 September to 2 October 2026).