"Did the ads work?" sounds like a simple question, but on TikTok Shop it is surprisingly tricky. A shopper might see an ad, then a creator video, then buy during a LIVE. Ads Manager may count that sale, the affiliate report may count it and Seller Center counts it once as an order. Without understanding attribution, brands either overestimate ad performance and overspend, or underestimate it and cut ads that were working.
This guide explains how attribution for TikTok Shop ads generally works and how to report sensibly. Attribution models, windows and metric definitions change and vary by campaign type and market, so check the definitions shown in Ads Manager for your campaigns.
What Attribution Means
Attribution is the rule that decides which marketing touch gets credit for a sale. For ads, it usually means: if a shopper interacted with an ad and then bought within a set time, the ad gets credit. The time limit is the attribution window, and the type of interaction, a click or a view, matters.
Click-Through and View-Through
- Click-through attribution: the shopper clicked or tapped the ad, then purchased within the click window.
- View-through attribution: the shopper watched the ad without clicking, then purchased within the view window.
View-through credit is more generous, because many people who watch an ad would have bought anyway. Check which types are included in your reports and treat view-through numbers with more caution.
Attribution Windows
Windows define how long after an interaction a purchase still counts, for example a number of days after a click and a shorter period after a view. Longer windows credit more sales to ads. When comparing campaigns or periods, make sure the same windows are used, or the comparison is meaningless.
Shop Ads Reporting
Shop ad campaigns, including automated formats such as GMV Max, often report gross revenue, orders, cost per order and return on ad spend directly from TikTok Shop sales. Some automated campaigns may count sales from a wider range of surfaces, including organic content, once the campaign is running. Read the campaign's metric definitions to know exactly what is counted. Our guide to ads and GMV Max explains campaign types.
ROAS and Its Limits
Return on ad spend (ROAS) is attributed revenue divided by ad spend. A ROAS of 4 means four units of revenue per unit spent. But ROAS uses revenue, not profit. A product with thin margins can lose money at a ROAS that looks healthy. Calculate break-even ROAS for each product: the ROAS at which gross profit after fees, shipping, discounts and commissions equals ad spend. Only ROAS above break-even is profitable.
Calculating Break-Even ROAS
Start with the selling price after discounts. Subtract product cost, platform fees, shipping, packaging and any affiliate commission. What remains is the gross profit per order before ads. Break-even ROAS equals the selling price divided by that gross profit. For example, if a product sells for 40 and leaves 12 of gross profit, break-even ROAS is 40 divided by 12, or about 3.3. Any campaign below 3.3 loses money on each attributed order, unless those customers come back to buy again. Recalculate whenever prices, costs or commissions change.
Overlap With Organic and Creator Sales
The same order can appear in several reports. An ad may get credit in Ads Manager, a creator may get commission for the same order if the shopper used their link, and the order appears once in Seller Center. Adding Ads Manager revenue to affiliate revenue to organic revenue will overstate total sales. Always use Seller Center as the source of truth for total orders and revenue, and treat channel reports as overlapping views.
Reconciling Ads With Seller Center
- Export total shop revenue and orders from Seller Center for the period.
- Export ad-attributed revenue and spend from Ads Manager with the same dates and time zone.
- Compare attributed revenue to total revenue to see the share credited to ads.
- Check whether the share is realistic given your organic and creator activity.
- Track total revenue and blended efficiency over time, not just attributed ROAS.
Our guide to data exports and dashboards shows how to combine sources.
Blended Metrics
Blended metrics, such as total ad spend divided by total shop revenue, sometimes called marketing efficiency ratio, avoid attribution arguments. If you increase ad spend and total revenue rises proportionally, ads are probably working. If spend rises but total revenue stays flat, ads may be claiming sales that would have happened anyway. Watch blended metrics alongside attributed ones.
Incrementality Testing
The most reliable way to know whether ads drive extra sales is to test. Approaches include pausing ads for a product for a period and measuring the change in total sales, running ads in one market but not a similar one, or using lift studies where available. Tests are imperfect, but they tell you more than attribution reports alone.
Time Zones, Currencies and Refunds
Mismatched time zones between Ads Manager and Seller Center shift sales between days. Currency differences complicate comparisons. Ads reports usually show gross revenue at order time and may not subtract later cancellations and refunds. Account for these when calculating true performance.
Reporting Ads to Stakeholders
When reporting ad performance to founders, investors or clients, show both views side by side: attributed results from Ads Manager and blended results from total shop data. Explain the attribution settings in one line and note major events, such as a viral creator video, that may have inflated or deflated attributed numbers. Honest reporting builds trust and leads to better budget decisions than presenting the most flattering number.
A Worked Example: Two Views of the Same Month
A brand spends a set budget on shop ads and Ads Manager reports a ROAS of 5. The team checks Seller Center and finds that ad-attributed revenue equals 70 percent of total shop revenue, even though creators posted heavily that month. They calculate break-even ROAS at 3.2 for their main product, so the reported ROAS is profitable. To test incrementality, they pause ads on one product for two weeks; total sales of that product drop by about 40 percent. The brand concludes ads are driving real sales, though less than the attributed figure suggests, and sets targets using blended efficiency as well as ROAS.
Customer Lifetime Value and Attribution
Attribution reports usually judge ads on the first purchase only. If customers acquired through ads come back to buy again, the true value of those ads is higher than first-order ROAS suggests. Where your data allows, compare repeat purchase rates for customers whose first order was ad-attributed with those who arrived organically. For consumable products with strong repeat rates, it can make sense to accept a lower first-order ROAS. For one-off purchases, first-order profitability matters much more.
Common Attribution Mistakes
- Adding channel reports together and double counting.
- Judging ads on revenue ROAS without margins.
- Comparing reports with different windows or time zones.
- Ignoring view-through inflation.
- Ignoring refunds and cancellations.
- Never testing incrementality.
Attribution Reporting Checklist
- Know the attribution model and windows for each campaign.
- Calculate break-even ROAS by product.
- Use Seller Center as the source of truth for totals.
- Track blended efficiency.
- Align time zones and dates.
- Adjust for refunds.
- Run periodic incrementality tests.
Desnora builds clear reporting for TikTok Shop advertising. Explore our TikTok Shop services or get in touch.
Frequently Asked Questions
What is attribution in TikTok ads?
The rule that decides which ad interaction gets credit for a sale, based on clicks, views and time windows.
What is an attribution window?
The period after an ad click or view during which a purchase is credited to the ad.
Why does Ads Manager show more sales than expected?
Ads may receive credit for sales also influenced by creators or organic content, and view-through credit can be generous.
What is ROAS?
Return on ad spend: attributed revenue divided by ad spend.
What is break-even ROAS?
The ROAS at which gross profit after all costs equals ad spend. Higher ROAS is profitable.
Can a sale count in both ads and affiliate reports?
Yes. Channel reports overlap, so use Seller Center for total sales.
What are blended metrics?
Metrics such as total ad spend divided by total revenue, which avoid attribution debates.
How do I test if ads are really working?
Run incrementality tests, such as pausing ads for a product and measuring changes in total sales.
Do ad reports subtract refunds?
Often not. Adjust for cancellations and refunds when calculating true performance.
Why do Ads Manager and Seller Center numbers differ?
Different attribution rules, time zones, currencies and refund treatment cause differences.
Should I trust view-through conversions?
Treat them with caution, as many viewers might have purchased anyway.
Which source is the truth for total sales?
Seller Center, which records each order once.
Should I accept a lower ROAS for repeat-purchase products?
Often yes. If ad-acquired customers buy again, their lifetime value can justify a lower first-order ROAS.
How do I report ad results honestly?
Show attributed results and blended results side by side, explain the attribution settings and note unusual events.