Run your Double 11 agency review within two weeks of the sale. Start with sales after returns and ad return by campaign. Check new customers and stock left. Agree three changes for 618 before the meeting ends. Keep unresolved refunds visible so the first review doesn’t become a premature sign-off on the final result.
Ask your China partner to bring the order export and the campaign spending file to the review. Finance should bring the settlement reconciliation. Use the printable review agenda to record the decisions while everyone is present.
The two-week timing is our recommendation. This review method uses source definitions checked on October 21, 2026.
Build the Double 11 agency review around one scorecard
Set the campaign’s start and end timestamps in China time. Keep the same group of orders throughout: orders paid inside that window, with subsequent refunds attached to those orders.
Put a refund cut-off on every export: the last date included in its refund data.
An all-store monthly refund total can include purchases from outside your campaign. Deduct partial refunds by value too. Call the result “sales after refunds” and let finance reconcile it to recognized revenue; tax treatment and recognition timing can differ.
Read the signs against your agreed plan and the same data definitions.
| Metric | Where the number comes from | Good sign | Bad sign |
|---|---|---|---|
| Sales after refunds | Paid orders matched to refund records | Retained sales meet the agreed plan | Paid sales rise while retained sales fall |
| Ad return by campaign | Ad spend and attributed orders, with the attribution window recorded | Refund treatment is stated beside the ratio | Different campaigns use different sales definitions |
| New customers | Buyer report matched to kept orders | New-to-store buyers are separated from repeat buyers | Followers or members are counted as buyers |
| Contribution | Net sales less product and selling costs | Campaign covers the agreed costs | Extra sales deepen the loss |
| Stock left | Warehouse export by SKU and condition | Sellable stock is separated from damaged returns | All returned units are counted as available |
| Cash received | Settlement file and bank receipts | Holds and pending payouts explain the difference | Net sales and bank receipts are treated as equal |
Ask the partner to open the underlying report during the meeting. Douyin’s own documentation separates the live sales display from deeper activity analysis.
Douyin’s October 2023 guide describes a real-time sales display and a shareable campaign report, alongside analysis of products and audiences. Source: Douyin E-commerce (抖音电商), October 2023. https://school.jinritemai.com/doudian/web/articlev0/aHvN4bfpmqMu
Request exports using the labels in your current account; the cited guide describes the tool in 2023.
Check what the ad-return number includes
Check the footnote beneath the ratio. In a response about Lanzhou Huanghe’s annual report, the company’s accountant reproduced its Douyin advertising calculation for 2025.
The filing reported advertising ROI of 1.05 and defined it as attributed gross merchandise value divided by ad spend. Its sales definition excluded the effect of later refunds and returns. Source: Beijing Xinghua Certified Public Accountants (北京兴华会计师事务所), response on Lanzhou Huanghe’s 2025 annual report, June 2026, printed page 56. https://static.cninfo.com.cn/finalpage/2026-06-13/1225369700.PDF
The 1.05 ratio applies to this company in 2025. In your review, show the dashboard’s ratio beside a calculation using the same orders after refunds. Record whether the attribution window includes sales after an ad click or view, and whether the report mixes paid and organic traffic. Reconcile overlapping campaign totals to unique store orders before presenting a combined result.
TheChinaPath calculator data, October 2026, supplies the starting values for this check. Our Douyin model starts with a RMB 250 order value and a 2.5 return on ad spend. These are editable planning defaults, not measured store results. The example assumes equal-value orders and a hypothetical 20% full-refund rate.
| Calculation | Before refunds | With 20% of orders fully refunded |
|---|---|---|
| Paid orders in the example | 100 | 100 |
| Orders kept | 100 | 80 |
| Sales retained (RMB) | 25,000 | 20,000 |
| Ad spend (RMB) | 10,000 | 10,000 |
| Sales retained / ad spend | 2.5 | 2.0 |
| Ad cost per kept order (RMB) | 100 | 125 |
Ad spend stays at RMB 10,000 after the refunds. Product cost and shipping still have to be paid. Add platform charges and the partner’s contracted fee before judging whether the campaign made a contribution to overhead. Use the invoiced amounts, including any fee credits after refunds.
Count advertising actually consumed, excluding unused wallet top-ups. Have finance distinguish recoverable returned inventory from write-offs and return-handling costs. Avoid deducting discounts again when they’re already in the paid sales amount.
In this example, compare the RMB 125 ad cost per kept order with the margin left before ads, after all other selling costs.
Count buyers who kept orders, then check stock
The sales reconciliation gives you the order list for the next check.
Ask the partner how the report defines a new customer. New to this store doesn’t automatically mean new to the brand across Tmall Global and Douyin. Keep those counts separate unless you can match buyers reliably within your permitted data access. If the available report can’t identify kept new buyers, record that limit and use the supported count without relabeling it.
Count buyers with at least one order they haven’t fully refunded; partial refunds still reduce sales value. Check their repeat purchases later, allowing enough time for the product’s normal buying cycle. A short-window repeat rate for a durable appliance says little about whether the acquisition worked.
Stock awaiting inspection belongs on a separate line from sellable units. For each SKU (stock-keeping unit), record the warehouse and any shelf-life limit before comparing quantities with ordinary-week sales and approving another purchase order.
Agree the changes before the next major sale
Use the review to prepare for 618, the June shopping festival. Name who will make each change and how you’ll check it. One change might be to stop ads for a variant until its return reason is fixed.
Another could make stock confirmation a condition for raising campaign spend. The warehouse owner would confirm sellable units against every promoted SKU before the agency receives approval to increase spend.
Choose three changes in total, ranked by their cost to the business.
Write the budget effect and due date in the printable’s decisions box. The brand should assign its own approvals alongside the agency’s work, especially when a late product decision caused the problem.
Compare delivery with the agreed campaign plan. If you use last year’s sales, match the campaign length and the time allowed for refunds. Explain any change in assortment.
If the review exposes a contract gap, check the Tmall Partner contract terms before changing the commission base or handover duties. An operational meeting doesn’t itself rewrite the signed agreement.
The cross-border ecommerce service scope can help define which store duties the next agreement needs to cover.
For a renewal, use the Tmall Partner guide or the Douyin Partner guide to check channel responsibilities. Compass provides the partner-search route if a replacement is justified.
Questions before the review
Should we wait until every return is closed?
Meet within two weeks, while the team can still explain what happened. Sales stay provisional while refunds remain unresolved. Set a second reconciliation date based on the orders and after-sales cases still pending. Keep a dated record of the change between the two readings.
What is a good Double 11 return on ad spend?
Use the campaign’s cost structure to judge it. The same sales-to-ad-spend ratio leaves different amounts after refunds and product costs. Agree which costs the review includes before comparing partners. A target borrowed from another category can reward sales that leave your brand with a loss.
Should a weak Double 11 trigger an agency change?
Start with who controlled the failure. A stock shortage caused by the brand requires a different fix from an agency spending beyond its agreed limit. Compare the evidence with the signed scope and targets. If you decide to replace the partner, document access and handover responsibilities before setting the cutover date.
Request a Compass shortlist based on the gaps your review has documented.
Updated October 22, 2026
Working on China? Let's grab a coffee.
Tell us where you are now and where you want to get to.
