Baymard Institute’s running average of documented studies puts online cart abandonment at about 70% (Baymard Institute). Some of that is window shopping, and Baymard is clear that much of it is unavoidable. But a visitor who picked a product, put it in a cart, and got as far as typing an email address told you exactly what they want. For most small stores, that is the highest-intent traffic they ever get, and the usual response is silence.
Why manual follow-up does not scale
Nobody can email forty abandoned carts by hand on a Tuesday. So the work either does not happen or happens in a weekly batch, days after the shopper has bought from someone else. The constraint is not intent. It is that outreach depends on a person noticing, and traffic grows faster than attention.
One event, one sequence
The trigger is a single event: a cart with an email address attached that has not checked out. Automation owns everything after that.
- Wait a short interval, so you do not email someone who is still deciding.
- Check consent and frequency caps before anything sends.
- Respect quiet hours. A recovery message at 3 a.m. is a worse brand signal than no message.
- Send a first reminder about an hour after abandonment (H+1), a second at 24 hours (H+24), and a last one on day 3 (D+3) only if the first two were ignored. Each shows the exact items left behind.
- Stop the moment an order arrives.
The principle: the highest-intent moment in your funnel deserves a response, on a schedule, whether or not someone remembers. The tool is an email platform with store integration, such as Brevo or Klaviyo, which can fire a workflow on a cart event and exit it the moment a purchase is recorded.
Gen AI helps at the edges: writing the reminder around the specific products, and deciding whether a nudge is enough or a small incentive is justified. Most recoveries do not need a discount, and handing one to every shopper trains your best customers to abandon on purpose.
What good looks like, with a caveat
Bloomreach reports that the Swiss online retailer Flaschenpost saw a 42% conversion rate among customers who engaged with its automated cart recovery campaign, with an average order value of CHF 294 on recovered carts (Bloomreach case study). Two cautions. The figure covers customers who engaged, not every abandoned cart. And it comes from the vendor that sells the tool. Read it as proof the mechanism works, not as a rate to expect.
Measure it your way
Run the sequence for 30 days on half your traffic and hold the other half back. Compare recovered revenue, unsubscribe rate, and refund rate. The holdout is what tells you whether the sequence created sales or only reached people who were going to buy anyway.
Boundaries
Automation owns the trigger, the timing, the consent check, and the stop rule. Gen AI owns the wording. Humans own the discount policy and anything touching customer data rules such as CASL consent in Canada.
The cart is not abandoned. Your follow-up is. Where in your funnel does a shopper with their hand on the wallet hear nothing from you?