How do you turn Black Friday buyers into repeat customers?

Start with the first-time buyers from your Black Friday sale. Check the first order and whether the buyer can receive marketing, then give them a reason to come back that fits what they bought. Track how many place a valid second order and what those orders contribute after discounts and returns. Do not assume every sale buyer needs another coupon.

A Black Friday order is an acquisition result. It is not yet a repeat relationship. The useful question after the sale is narrower than “What should our post-purchase flow contain?”: which first-time buyers from this promotion have a credible next purchase, and did the follow-up add value?

This guide deals with buyers whose first order came during the store’s Black Friday/Cyber Monday promotion. It works whether the promotion used a code, an automatic discount or sale pricing. The post-purchase flow guide covers what belongs in a general flow; the work here is to define and evaluate the BFCM group within it.

Who counts as a first-time Black Friday buyer?

Fix the cohort before looking at its results. Use the dates of your promotion, find customers whose first valid order fell inside them, and record the first-order product, promotion and order status. A prior customer who bought again during BFCM is a repeat buyer already; mixing them into the acquisition cohort answers a different question from the one this page asks.

Shopify’s customer cohort report groups customers by first-order date and shows their later purchases. Its documented first-order filters include sales channel, marketing channel and product, and it compares reporting-period cohorts. To fix the group to your exact BFCM dates and offer, export orders, keep one customer per first order and preserve the order identifier so the same buyer is not counted twice.

Mark how the first purchase was promoted. A code is only one method. Shopify’s Sales by discount codes report also groups automatic discounts by name, but an order with combinable discounts may appear in more than one row. A sale price that was never recorded as a discount has no discount name for that report to group. Do not add the report’s rows and call the sum unique buyers. Join the promotion details to unique orders, or keep an “unknown promotion” group when the record cannot answer it.

The first-order product also changes what a sensible return visit looks like. A buyer may have bought a gift rather than something they will use. If you cannot tell, do not write as if they personally need a refill. That is an applicability check, not a prediction about how many BFCM orders are gifts.

What should happen before the next offer?

Before any second-order offer, confirm the first order has been handled, the buyer can currently receive marketing, and they have not already placed a second order. Check fulfillment and any return or refund problem before placing a promotional message behind it. Shopify’s store notifications are triggered by order, fulfillment, refund and exchange events; keep those service updates functioning as service updates.

Next, check marketing eligibility. Having an order email address is not the same as permission to send promotions. Shopify says promotional content should go to customers who agreed to receive marketing. Put the current subscription state in the send check, not merely in the original BFCM export.

Finally, remove buyers who have already ordered again from a second-purchase prompt. If your store uses Klaviyo, a Placed Order-triggered flow can narrow entry by the items in that order, and profile filters are checked again before each flow action. That gives you a way to skip a later message when the buyer no longer qualifies. Verify the actual logic with sample orders, including an order made between the first and second messages. For another platform, check its own entry and pre-send filter behavior.

Should you offer another discount for the second order?

Not by default. First check whether there is a credible next purchase. If you test another discount, compare it with a version without the extra discount and judge contribution after its cost.

The BFCM basket might have been a gift or a stock-up purchase. Neither automatically calls for a reorder prompt immediately after the sale. If the store cannot tell who will use the item, avoid copy that assumes the buyer will. If there is no credible next purchase yet, leave the buyer out of the second-order offer. The replenishment guide covers a separate decision for products with a repeat-use cycle.

A first discount alone does not tell you whether that buyer will return. In three field experiments reported in Marketing Science in 2004, deeper current discounts increased future purchases among first-time customers of a direct-mail durable-goods catalog, while reducing future purchases among its established customers. Those results do not estimate what a 2026 ecommerce BFCM discount will do at your store. They do rule out treating “discount buyer” as a proven synonym for “one-time buyer.”

For a second-order incentive, write down the alternative before sending it. Compare a relevant follow-up without the extra discount against the version with it, or reserve an eligible no-message group if the question is whether the follow-up itself adds value. Judge the extra orders and the contribution after the incentive’s cost. A second coupon that simply reduces the price of an order the buyer would have placed anyway is not a retention win.

How do you know whether BFCM buyers came back?

Use a fixed cohort and a declared window. Count customers in the first-time BFCM cohort who place at least one valid second order within that window, then divide by the number of customers in that same cohort. Choose the window based on what the store sells and allow enough follow-up time for every cohort member; do not compare a mature cohort with one that has barely had time to reorder. Record how refunds, cancellations, subscriptions and split orders are handled.

For example, suppose 400 eligible first-time BFCM buyers are followed for the same declared window and 80 distinct buyers place a valid second order. Their repeat-buyer share is 80 ÷ 400 = 20%. Eighty orders are not necessarily 80 repeat buyers. If you report the outcome for all first-time BFCM buyers instead, use that whole cohort as the denominator and label it separately from the marketable subgroup. See repeat purchase rate for the other denominator choices this metric can hide.

The same cohort report includes repeat orders in period 0, the same reporting period as the first order. Decide whether an order made during the BFCM sale itself answers your question or whether you want the post-sale second order. State the choice. For revenue, the sales reports define net sales after discounts and sales reversals, which include returns and cancellations. Add the store’s variable product, payment and fulfillment costs before calling the second orders profitable.

An observed second order is not proof that the email produced it. Buyers can return without a message, and a platform’s attributed-sales report answers a different question. If you need to estimate the effect of the follow-up, randomly hold out eligible buyers from that promotional treatment, keep the groups comparable, and compare their second-order contribution over the same window. The email-sales tracking guide explains why a credited order and an added order are different measures.

The useful BFCM result is a cohort you can account for: who bought for the first time, who could be contacted, who bought again after the sale, and what remained after the next order’s discount and costs. That record tells you which follow-up to keep for the next promotion without assuming the first discount determined the customer’s future.