Cart recovery in a sale: the timing that works when everyone is messaging

Sale-season abandonment is different in kind, not just in volume. Shoppers open five stores, add to five carts and buy from one. A recovery sequence that works in a quiet month — a message after an hour, another the next day — is competing during a sale with dozens of identical messages, and a shopper who has already bought elsewhere. Timing, content and restraint all change.

What changes in a sale

• Decisions are made in minutes, not hours. A message sent an hour later often arrives after the shopper has bought from a competitor.
• Price is the objection. Most sale carts are abandoned to compare, so the message needs to answer the comparison, not just remind.
• Stock is the urgency. 'Only a few left in your size' is true in a sale and worth saying; in a quiet month it reads as a trick.
• Everyone is sending. Frequency that felt attentive in August feels like spam in October — send fewer, better messages.

A sale-season sequence

• 15–20 minutes, WhatsApp. The cart contents, the sale price, and a one-tap resume link that opens the cart prefilled and OTP-bypassed. No discount yet.
• 3 hours, WhatsApp or SMS. Stock and time urgency if either is genuine — the sale ends, the size is low. Still no extra discount.
• Next morning, email. The comeback offer, if you use one: a small extra on prepaid only, so the recovered order is also a safe one.
• Stop the moment they buy. From you or anywhere — a shopper who has completed the purchase and still gets a reminder is an unsubscribe.

Reach only the shoppers you can actually reach

None of this works on anonymous sessions. Recovery in a sale depends on capturing a verified phone number at the very start of checkout — before the address form, where most Indian carts die. A phone-first, OTP-verified checkout turns an abandoned cart into a contactable person; anything else is a guess at an email that bounces.

Template approvals: do it now

WhatsApp business templates need approval before they can be sent, and approval queues lengthen in the run-up to festive season. Write the sale variants of your recovery templates in September — with variables for cart value, sale price and resume link — and get them approved while the queue is short. A template rejected on the first day of the sale is a sequence that never runs.

Measure what came back

Attribute recovered revenue per message, not per sequence. In a sale the first WhatsApp usually earns most of it and the email earns the least; knowing that lets you cut the noise next year. If the numbers show the sequence is mostly recovering orders that would have completed anyway, shorten it — restraint is a feature during a sale.