Should Refunded Shopify Orders Count in a Discovery Survey?

Keep a refunded buyer’s answer to “Where did you first hear about us?” The refund changes how much of that order’s value remains. It doesn’t change what the buyer told you about discovering the store. For a channel report, keep two columns: the number of answering orders and the amount left on those orders after refunds.

A Shopify merchant comparing export fields found the same amount under a discount and a refund and asked how to avoid counting it twice. A survey report adds another place to make that mistake: a buyer who answers a discovery question and a follow-up can produce two CSV rows for the same order.

One creator mention, even if the order is refunded

Suppose you ran a discovery survey on three fully paid USD orders. All three buyers answered. The figures below are illustrative, and the order totals already reflect any checkout discounts.

Two of the three answering orders mentioned a creator. As of this export, those two orders account for $45 of the $95 left after refunds. Dropping order A would change the discovery count from two out of three to one out of two. Charging the full $80 to the creator channel after it was refunded would make the opposite mistake.

If the buyer on order B also named the creator in a follow-up, keep the name alongside the discovery answer. The $45 belongs to order B once. Neither the mention nor the dollar amount shows whether this customer would have bought without the creator.

Build a worksheet you can update next month

1. Pick the orders and an as-of date. For example, use orders placed in October and label the refund column “checked November 13.” Start with paid orders in one currency for this calculation. Keep unpaid orders, edits you can’t reconcile and missing matches in a review group.
2. Keep one discovery answer per order. In OrderSurvey’s Responses CSV, filter to the exact question you asked. Keep any follow-up in a separate column. Flag conflicting answers for the same order before assigning it to a channel. How to read the survey export
3. Match Shopify’s order ID. Export the same order period from Shopify. Use its Id column and the survey’s Order ID as text. If the survey uses a Shopify Order ID such as gid://shopify/Order/12345, match its numeric part to 12345. The displayed order number, such as #1042, is a different field. Shopify can put additional items from one order on separate lines. Keep one financial row per order and spot-check a match in admin. Shopify’s order export fields
4. Calculate once, then group. For a fully paid order in this worksheet, subtract Refunded Amount from Total. Check a partial refund against the order details. A checkout discount already reflected in Total needs no second deduction. Keep orders without survey answers in an “unanswered” group if you’re comparing with the whole store.

What happens when a refund arrives later?

If order B receives another $45 refund in December, its amount left becomes $0. The October buyers still gave two creator answers. Save the November worksheet before updating it so you can explain why the dollar figure changed. To compare two months’ buyers fairly, give both groups the same refund observation period.

This worksheet groups buyers by the date they ordered. Shopify’s sales reports can record a return in the period when it was processed, and order totals can include shipping and taxes. Use Shopify’s report definitions when you need net sales, and the payments report when you need to reconcile money refunded. Shopify’s explanation of sales discrepancies

To collect the discovery answer on Shopify’s Thank you and Order status pages, install OrderSurvey. Free includes 100 responses per rolling 30 days. The CSV export used for this worksheet is included in Unlimited at $24 USD per month. See plans and features

Frequently asked questions