"We have cart recovery covered, it's in Klaviyo" is one of the most common things a store owner says right before finding out what that flow actually catches. The flow runs. The revenue it recovers is smaller than the sequence's completion rate would suggest, and the gap between those two numbers is where this false belief lives.
A flow and a conversation solve different problems. Confusing one for the other is the mistake, not the flow itself.
A cart-abandonment flow in Klaviyo fires on a timer: an email an hour after the cart sits idle, another the next day, a discount code on day three. It's the same sequence every time, regardless of who the shopper is or what stopped her from checking out. She might have gotten distracted by a phone call. She might have hit a shipping cost she wasn't expecting. She might have had a question about sizing that nobody answered. The flow doesn't know which, and it doesn't ask.
That's not a criticism of the flow. It was built to do exactly this: reach a list at scale, on a predictable schedule, without needing a human to read anything back. It's good at that job. The problem shows up when a store treats "the flow is running" as equivalent to "cart recovery is handled," because those are not the same claim.
A shopper who abandons a cart because of a real, answerable question, sizing, shipping cost, return policy, doesn't get that question answered by a flow. She either finds the answer somewhere else, on her own, or she doesn't buy. The flow keeps running its sequence either way, unaware that the actual obstacle was something a two-minute conversation could have resolved.
A real conversation catches that moment differently. A shopper who texts back "does this run small" gets an answer to that specific question, not a generic follow-up written for an audience of thousands. The reply changes what happens next, because the message that comes after it responds to what she actually asked instead of moving to the next scheduled step regardless of what she said.
That's the whole distinction. A flow completes on schedule. A conversation completes when the shopper's actual question gets answered.
Open rates on abandoned-cart emails look fine. Completion rates on the sequence look fine. Every dashboard a store checks says the flow is doing its job, because those dashboards measure whether the flow ran, not whether it resolved what actually stood between the shopper and checkout. A flow can run at 100% completion and still miss every shopper whose reason for abandoning was a question, not indifference.
This is why the false belief is so sticky. Nothing about the flow's own reporting flags the gap. A store owner has to go looking for it, usually by noticing how few shoppers ever reply to anything, before the gap becomes visible at all.
None of this means the flow should be turned off. Email still reaches shoppers SMS can't, and a scheduled sequence still recovers carts abandoned for reasons a conversation wouldn't change anyway, someone who just changed her mind, someone who was never going to buy that day regardless of what she was asked. The flow keeps its job.
What changes is what a store assumes that job covers. "The flow is running" answers one question. "Every recoverable cart is being recovered" is a different, larger claim, and the gap between the two is exactly the shopper who had a real question and never got it answered.
Run your next 250 abandoned carts through CART, $0 charged, and see what it recovers before you commit to anything.
Book a Revenue Proof Call →