Most TikTok Shop samples never sell. When should a sample be free, and when refundable?

KEY TAKEAWAYS

1. Most samples produce no video, and a minority of creators produce almost all the sales. Budget for that.
2. The rising sales per video in the new report is mostly selection. Use the first video to find sellers, not to push volume.
3. Price a sample in orders. Break-even orders per sample is the loaded sample cost divided by the profit per order.
4. Refundable samples turn sample cost into a cost per seller. Free samples win only if switching would lose a large share of your sellers.

On 14 October 2026 Digiday reported a Flywheel study built on data from Reacher, an affiliate tool for TikTok Shop. It covered about 1.5 million product samples sent to creators. The headline was simple: creators who post many videos sell far more per video, so brands should seed widely and push for volume. Flywheel published its own summary on 19 October.

The data is useful. The headline reading is not quite right. This piece covers what the samples show, why the volume curve is mostly selection, and how to decide whether each sample should be free or refundable.

What do 1.5 million samples actually show?

Reacher's published data page is more careful than the coverage. For samples approved between 17 February and 1 July 2026, it reports:

• 46.57% of shipments from most brands produced at least one video. Flywheel's own brands reached 51.84%.
• Across all mature shipments, the average was 0.63 to 0.70 videos per sample.
• Only 3% to 4% of shipments produced three or more videos. Reacher's two checks disagree on both figures.
• 8.66% of video records generated at least one sale.

That measured rate sits well below the planning tables some agencies publish. MediaLabs lists post rates of 65% to 85% for briefed, retained creator programs, without saying they were observed. Reacher also counts video-attributed sales. In LIVE-heavy categories that misses a lot: FastMoss estimated September 2026 collectibles sales at 31.3% LIVE and 1.8% video.

Sales are concentrated. In Reacher's snapshot of its creator roster, the top 1% of creators held 90.29% of recorded sales. In the same snapshot, 85.86% of creators showed no sign of any sale.

One correction matters here. Digiday's summary listed 14% of creators generating sales next to the sample figures, and some coverage read it as the share of sampled creators who sell. Reacher's page says the 14.14% figure comes from a profile snapshot of the creators it tracks, not specifically creators who received samples. The share of your sample recipients who sell is something you have to measure yourself.

Why do sales per video climb with volume?

The table groups each creator and brand pair by how many videos the creator posted.

Coverage read this as compounding. Post more and each video earns more. Flywheel's CEO, Deren Baker, told Digiday it gives you "more shots on goal."

Reacher's own notes describe the table as observational and shaped by survivorship. That is the better reading. A creator whose first video sells has every reason to post again. A creator whose first video sells nothing usually stops. The 20-plus group is not a sample of average creators who kept going. It is weighted heavily toward creators whose content already worked.

Reacher's earlier report shows the filter at work. In a same-brand sequence, 5.81 million first videos averaged $24.61 in sales. By the tenth video, 394,845 videos remained, about 7% of the starting count, and they averaged $44.94. The share of videos with any sale rose only from 7.6% to 10.1%. Most of the change is who is still posting, not what posting did to them.

Affiverse's coverage makes the same point.

Does paying creators to post more raise sales?

This is the closest the data gets to a test. Reacher compared the same creator and brand pairs before and during a paid retainer, across 156 pairs in 22 shops. Videos rose 49.19%. Sales rose 13.41%. A second comparison, retainer pairs against a proxy group without direct pay, 99 pairs in 14 shops, found retainer pairs with 42.51% lower sales. Reacher's own conclusion is that no universal uplift is supported.

Across shops, videos per sample barely tracked sales. The correlation was -0.003.

So the two readings agree on the first half of the advice. Seed broadly, then keep the creators who sell. They split on the second half. Paying for more volume from creators who have not sold buys videos, not much revenue. TikTok's own matchmaking playbook leans the same way. It advises matching products to creators on purpose, not sending many products to every creator.

What does a sample have to earn?

Treat each sample as a purchase of orders. The break-even is one division:

Break-even orders per sample = loaded sample cost / profit per order

Loaded sample cost covers the product, outbound shipping and packing. Profit per order is price minus product cost, fulfilment, the creator's commission and TikTok's referral fee. Sources disagree on that fee. FastMoss read TikTok's table in June 2026 as 6% for most US categories. beBOLD reported in September that seller guides saw 8% from 4 August, while the public table still showed 6%. Webgility, writing on 22 October, advises modelling both. Check your own Seller Center.

TikTok uses the same unit. Its Auto-optimized Samples feature, open to managed US sellers, defines ROI as GMV divided by average order value. That is orders per sample, not return on cost. Sellers can lower the target from TikTok's estimate in steps of 0.25. A lower target brings more content and fewer orders.

Free or refundable?

TikTok's US rules make the two routes behave very differently.

A refundable sample that never earns a refund is just a sale. That turns sample cost into a cost per selling creator. The price is friction. The creator pays first, and some will not. TikTok's EU seller page reports one data point. Raising the goal may cut the chance a creator buys by about 10%, based on UK data from November 2024 to March 2025. Nobody has published the gap between free and refundable take-up. One cost changed recently: TikTok's US guide says that from 1 September 2026, commission applies to refundable sample orders placed through an affiliate link. The older terms pages say no commission is paid on those orders.

Hubfluence suggests refundable samples for expensive products. There is a tension in that advice. The upfront cost is the friction, and it is larger on an expensive item. The formula below works around it with your own numbers rather than price bands.

Free beats refundable only if switching loses more profit from sellers who drop out than it saves on samples sent to creators who never sell. As a share of your sellers, the line is:

(1 - s) x C / (s x V)

Here s is the share of sample recipients who sell, C is loaded sample cost and V is profit per selling creator.

A worked example (hypothetical)

A brand sells a $30 product. Product cost is $8, fulfilment $5, commission 15% ($4.50) and an assumed 8% referral fee ($2.40). Profit per order is about $10. A sample costs $13 to send, so break-even is 1.3 orders per sample.

It sends 100 free samples for $1,300. Assume 15 creators sell and the batch brings 2 orders per sample, close to the $63 in sales per sample Reacher reported for the 90% of shops outside its top tenth. That is 200 orders and about $2,000 in profit, or about $133 per seller. The batch pays.

With refundable samples, the 85 creators who never sell cost nothing, which saves about $1,105. The line is 0.85 x 13 / (0.15 x 133), about 55%. Free sampling only wins if refundable samples would lose more than half of the selling creators. TikTok's 10% figure measures a different change, a higher goal, so test the switch on your own shop. A cheaper product or a higher seller share would lower the line.

Which route fits which creator?

TikTok's auto-approval rules can filter on 30-day sales, estimated posting rate, category and price, with a shop quota of 1 to 100. They do not apply to refundable or auto-optimized samples.

TikTok's Sample Integrity Policy gives creators one 7-day extension per free sample. After that the sample is overdue, and posting late is allowed but may lower the creator's posting rate, which sellers see as Ok, Good or Better. An overdue sample also keeps using the creator's sample limit until they post, so some late videos still arrive.

A seven-step sampling method

1. Load the sample cost. Product, shipping, packing and staff time per sample.
2. Work out profit per order. Confirm your referral fee and commission in Seller Center.
3. Set break-even orders per sample. Divide step 1 by step 2.
4. Run the first wave as a cohort. Judge it after about 50 days, allowing 14 days to post, one possible 7-day extension and about 30 days for returns to clear. Record post rate, seller share and orders per sample.
5. Rank creators by orders, not videos or followers. Reacher found small creators earn more per follower and large ones more per video, and a Journal of Marketing study of paid Instagram endorsements found return on spend over three times higher for nano-influencers. Keep sellers, give them new products first, and stop sampling non-posters. Test paid retainers on proven sellers against a holdout, since the evidence is mixed. Sizing that kind of test is covered in how many concepts a budget can test.
6. Switch open requests to refundable if the cohort falls below break-even, or if the formula's line sits above the share of sellers you expect to lose.
7. Brief for disclosure. The FTC treats a free product as a material connection. It wants the disclosure in the video, not only the caption, and it expects brands that send products to ask for it and monitor posts.

Winning creator videos often deserve paid reach next. Our piece on creator and brand handles covers who should run them. Check the soundtrack first, as music rights in creator content explains.

The short version

Most samples produce no video, and a small group of creators makes most of the sales. The new volume curve shows which creators survive, not what volume does. Price every sample in orders, send free samples where one or two orders repay them, and let refundable samples screen the creators you do not know. If you want help reading creator and channel data as part of your wider visibility work, join the waitlist.

Questions

What share of TikTok Shop samples turn into a video?

In Reacher's data on samples approved between February and July 2026, 46.57% of shipments from most brands produced at least one video, and Flywheel's brands reached 51.84%. Across all mature shipments the average was 0.63 to 0.70 videos per sample, depending on which of Reacher's checks you use, and only 3% to 4% produced three or more.

Does posting more videos make a creator sell more?

The data cannot show that. Creators who post 20 or more videos average $74.07 per video against $7.19 for one-video creators, but Reacher's own notes flag it as shaped by survivorship. Creators whose first videos sell are the ones who keep posting. When brands paid retainers to the same creators, videos rose 49% and sales 13%.

How do TikTok Shop refundable samples work?

The creator buys the sample, then gets a full refund from the seller once their content drives the seller's sales goal, one sale by default, within 120 days. Orders count only after the return period, typically 30 days. If the creator never hits the goal, the sample stays a normal purchase.

Do creators need to disclose free samples?

Yes, in the US. The FTC treats a free product as a material connection, wants the disclosure in the video itself, and says a platform's disclosure tool may not be enough. A brand that sends free products should ask for disclosure and monitor the posts.

Sources

1. Digiday, Future of TV Briefing. TikTok Shop success = small creators + more videos digiday.com
2. Flywheel, Your TikTok Shop sampling strategy may be wasting your budget flywheeldigital.com
3. Reacher x Flywheel, TikTok Shop whitepaper data reacher-flywheel-whitepaper-data.vercel.app
4. Reacher x Flywheel, TikTok Shop data report V2 reacherflywheelexternal20260708.vercel.app
5. TikTok Shop Academy US, Guide to samples. Free and refundable seller-us.tiktok.com
6. TikTok Shop Academy US, How to set up and manage samples seller-us.tiktok.com
7. TikTok Shop Academy US, Refundable samples terms and conditions for sellers seller-us.tiktok.com
8. TikTok Shop Academy US, Refundable samples terms and conditions for creators seller-us.tiktok.com
9. TikTok Shop Academy EU, Refundable samples for sellers seller-ie.tiktok.com
10. TikTok Shop Academy US, Auto-optimized samples seller-us.tiktok.com
11. TikTok Shop Academy US, How to set up sample auto-approval automations seller-us.tiktok.com
12. TikTok Shop Academy US, 8 tips for successful seller and creator matchmaking seller-us.tiktok.com
13. FTC, Disclosures 101 for social media influencers ftc.gov
14. FTC, FTC's Endorsement Guides. What people are asking ftc.gov
15. Affiverse, TikTok Shop research puts repeat creator content and shopper trust in focus affiversemedia.com
16. FastMoss, TikTok Shop seller costs in the US 2026 fastmoss.com
17. beBOLD, TikTok Shop 8 percent referral fee bebolddigital.com
18. Hubfluence, TikTok Shop refundable samples explained hubfluence.io
19. TikTok Shop Academy US, Sample integrity policy seller-us.tiktok.com
20. FastMoss, US TikTok Shop statistics. September 2026 sales data fastmoss.com
21. MediaLabs, TikTok Shop performance benchmarks 2026 medialabs-co.com
22. American Marketing Association, Influencer marketing ROI research shows influencers with smaller followings are more cost effective ama.org
23. Webgility, TikTok Shop Black Friday 2026. Fees, discounts and margins webgility.com

NEXT STEP

See where your brand stands.

Your Digital Authority Score, who AI recommends for your target questions, and the placements that close each gap. Your first report is free.

Get your free report →