An abandoned cart flow has one job: catch the people who intended to buy and got interrupted. It is good at that and useless at everything else, and knowing the difference saves you from writing better emails to solve a product page problem.
The sequence
Hour 1 — the reminder. Short, the product shown, a direct link back to the cart. No discount. Most of what this flow recovers is recovered here, by people who genuinely forgot.
Hour 20 — the objection. Address the thing that most often stops people in your category: delivery time, returns, fit, whether it works. One objection, answered properly.
Day 3 — the last call. Scarcity if it is real. A discount only if you have decided that is a trade you want to make permanently, because people learn.
Discount in message one and you have taught your best customers to put things in the cart and wait. The flow will look like it is performing while quietly costing you margin on orders that were coming anyway.
What it cannot recover
Someone who left because shipping doubled the total did not forget. They decided. Sending them the same total again changes nothing.
Same for a missing payment method, a delivery date too far out, or a forced account. Every one of those is fixed on the product page or in the checkout, not in an email.
The upstream fixes worth more than the flow
Show delivery cost before the decision. Give a date rather than a window. Offer guest checkout. Show accepted payment methods early.
Each of those prevents abandonment rather than chasing it, and prevention does not need a send.
Measuring it honestly
Attribution in this flow flatters itself badly. Some proportion of recovered carts would have returned anyway — people get interrupted and come back without help.
If your volume allows it, hold out a small percentage and compare. The real number is almost always lower than the dashboard says, and still worth having.
Talk it through
Send us your cart and checkout abandonment split.
Those two numbers tell us whether this is a flow problem or a product page problem, and they are usually the second.
Get a readA real read within one business day — not a sales call.
About the author
Manpreet Singh
Founder of Proscube. Leads the studio's Shopify, headless and CRO engineering, and works directly with founders rather than through account managers.
More about Manpreet