Skip to main content
The number of new-customer orders that had at least one refund in the period.

​
Formula

New Customer Refund Count = COUNT DISTINCT ( Orders with a refund in the period ) WHERE the refunded order is a new-customer order

​
Formula Components

​
How It Works

Customer type comes from the order that was refunded, not from the shopper’s status on the day of the refund. Upstack marks each customer’s earliest non-cancelled order as their new-customer order. A refund on that order counts here even if the shopper has placed three more orders since. A refund on any later order counts in Returning Customer Refund Count. Counting rules, as in Refund Count:
  • Each order counts once per period, however many partial refunds it had in that period.
  • Any refund Shopify records counts. Upstack doesn’t check the amount.
  • By default a refund is dated by when Shopify processed it. You can switch a dashboard view to the original order date from the dashboard options menu, under Refunds.
The date setting matters most for New Customer Refund %, which divides this count by new customer orders. On the default refund-date setting, the two sides can come from different months.

​
Example

An apparel brand acquires 3,600 new customers in November and 1,500 in December. Many November first orders are refunded in December. On refund date, December looks like a refund problem: 10% of new customers. Switch to original order date and the picture flips. November’s first orders had the higher refund rate, and 108 of their 216 refunds were processed in December, which is what inflated December’s view. December’s 4.0% isn’t final either, since some of its refunds will land in January.

​
How to Interpret It

Use refund date for operations, order date for acquisition quality. Refund date tells you how many refunds the team handled this month. Order date ties each refund back to the month and campaign that acquired the customer. A rising count with flat new-customer volume means first orders are failing more often: sizing, product expectations set by ads, or shipping damage. Read it next to New Customer Refunds to see whether the refunds are partial or full. Compare it with returning customers. Repeat buyers already know the product. When first orders refund at a much higher rate than repeat orders, look at what the ads and product pages promised, since that is the main thing repeat buyers no longer depend on.

​
FAQ

Yes. Any refund Shopify records on the order counts, and the order counts once per period no matter how many refunds it had.
No. Only refunds on the customer’s new-customer order count here. Refunds on later orders count as returning.
On the default refund-date setting, refunds from a big previous month land in a month with fewer new customers. Switch the view to original order date to tie refunds back to the month the customer was acquired.
No. It counts orders. An order with three refunded items counts once.
See all Adjustments metrics →