A tracked sale is useful. It is not the whole story.

A customer watches a creator, searches for the product later and buys on another device. Another customer uses a creator’s code after already deciding to purchase. Which sale did the campaign cause? A tracking system may record an action. It cannot automatically answer the counterfactual: what would have happened without the campaign?

Define what your report is actually counting

Attributed sales are sales assigned to a campaign under a chosen rule. Incremental sales are the additional sales that would not otherwise have happened. Those are not interchangeable terms. Be clear about the attribution window, the eligible action and whether the rule uses clicks, codes or another signal. Otherwise two reports can look contradictory while simply measuring different things.

Know where signals go missing—or overlap

• A customer may switch devices or decline tracking.
• A discount code may be shared beyond the creator’s audience.
• Several marketing channels may influence the same purchase.
• Returns, cancellations and duplicate events may distort gross sales.
• A view can introduce a product without receiving credit for the final purchase.

Agree on the rules before launch

Choose the campaign objective and document how results will be counted. Keep tracking consistent across the creators you want to compare. If you report attributed revenue, identify whether it includes taxes, delivery charges and refunded orders. Do not quietly change definitions halfway through and call the difference an improvement.

Use measurement to make a better decision

When feasible, a well-designed holdout or controlled test can help investigate incremental impact. It needs enough data and a sensible comparison, not just a before-and-after screenshot. For smaller campaigns, acknowledge uncertainty and combine sales signals with reach, creative quality and customer questions.

The common tracking methods, and their limits

• Unique links: show clicks and, with proper setup, purchases from that link. They miss people who see the video and search for the brand later.
• Discount codes: easy to measure, but codes get shared on coupon sites and can reward the code rather than the content.
• Platform metrics: views, watch time, saves and shares show attention, not sales.
• Post-purchase surveys: “Where did you hear about us?” captures influence links miss, but depends on memory.
• Sales lift in a period or region: shows the overall effect, but other factors can affect it too.

Decide what counts before the campaign starts

Agree on the main goal, the metric that represents it and the attribution window: how many days after seeing content a purchase still counts. Changing the definition after the results arrive is how teams end up arguing instead of learning. Write it in the brief so creators know what they are measured against.

Use more than one signal

No single method shows the full picture. Combining tracked links or codes with a post-purchase question and a look at overall sales during the campaign gives you a more honest view. If all three point in the same direction, you can be more confident. If they disagree, that is useful information too.

Read the numbers in context

• Was the product in stock and easy to buy during the campaign?
• Did a promotion, holiday or competitor launch happen at the same time?
• Did the content send people to a page that matched what they saw?
• Was paid promotion used on some pieces and not others?

Turn tracking into decisions

The goal is not a perfect report. It is a better next campaign. Use the data to decide which messages, formats and creators to continue with, what to change on the landing page and how to set budgets. Share results with creators too. When they see what worked, they can make better content next time.

We believe creator marketing should be accountable. Accountability includes admitting what a number cannot prove. A useful report is not the one with the biggest total. It is the one that tells you what was measured, what remains uncertain and what to test next.