Segment your abandoned cart emails by three customer types: first-time visitors who have never bought, one-time buyers, and repeat customers. Each group abandons for different reasons, so each gets a different message. First-timers need trust and reassurance, not a discount. One-time buyers respond to a reminder that references their previous order. Repeat customers usually just need a nudge — sending them a discount trains them to abandon on purpose. Most email platforms, including Omnisend, can split a cart flow on order count in a few clicks. This article shows exactly how to set the branches, delays and messages.

Why one generic cart email leaves money behind

If your store does €30,000–100,000 a month, your cart abandonment list is not one audience. It’s a mix of people who found you through a Meta ad ten minutes ago and customers who have ordered four times. Sending both the same “You forgot something!” email with a 10% code does two damaging things at once: it fails to answer the stranger’s real objection (can I trust this shop?), and it hands margin to a loyal customer who was probably going to buy anyway.

You feel this in the numbers. Cart emails get opened, some get clicked, revenue trickles in — but discount usage creeps up and repeat customers start “forgetting” carts every time they shop. That’s not recovery. That’s a self-inflicted price cut.

Why more ads and bigger discounts don’t fix it

The reflex fix is to push harder: raise ad spend to replace lost checkouts, or raise the discount from 10% to 15% so more carts convert. Both work briefly and both erode the business. Ad costs on Google and Meta keep climbing, so buying back a visitor who already reached your checkout is the most expensive possible way to close them. And a bigger blanket discount mostly transfers margin to people who didn’t need it. Neither approach touches the actual problem: the message doesn’t match the person.

Where the leak actually is: a quick calculation

Illustrative numbers, adjust to your store. Say you get 600 abandoned carts a month at an average cart value of €70. A generic flow recovering 6% brings back 36 orders, about €2,520. Now suppose 25% of those carts belong to existing customers. If a blanket 10% discount goes to all recovered orders, you gave up roughly €252 in margin — a chunk of it to repeat buyers who needed no incentive. Meanwhile, first-time visitors, who abandon most and trust you least, got a coupon instead of the shipping, returns and payment answers that would actually move them. A segmented flow attacks both ends: fewer wasted discounts, more relevant persuasion.

The three segments that matter (and the ones that don’t yet)

You don’t need ten branches. Start with three, based on purchase history:

1. Never purchased. Your biggest group by volume. They abandoned because of shipping costs, doubt about the product, or doubt about you. They need proof, policies and clarity.
2. Purchased once. They know the checkout works and the parcel arrived. Their barrier is lower; a personal-feeling reminder usually does it.
3. Purchased two or more times. Your regulars. Short reminder, no discount, warm tone. Protect margin here.

What you don’t need yet: splitting by traffic source, device, or product category. Those are refinements for later, once the three-way split runs cleanly. If you want to layer in cart value afterwards, personalizing abandoned cart emails by cart value is the natural next split.

The exact flow: triggers, branches, delays, messages

Here is the full specification, not a vague “set up segmentation”.

• Trigger: cart abandoned (checkout or cart updated, no order placed).
• Split condition: total number of orders — equals 0, equals 1, or is 2+. Evaluate at send time, not at trigger time, so a purchase mid-flow moves them out.
• Exit condition: order placed — the contact leaves the flow immediately, every branch.

Branch A: never purchased

• Email 1, 1 hour after abandonment: cart contents, one clear button back to checkout, plus your delivery time, return window and payment options stated plainly. Goal: remove doubt. No discount.
• Email 2, 24 hours: one or two short customer reviews of the abandoned product, answer to the most common pre-sale question. Goal: build trust.
• Email 3, 72 hours: last reminder; only here consider a small incentive like free shipping if your margin allows. Goal: close the fence-sitters.

Branch B: purchased once

• Email 1, 2 hours: “Your cart’s waiting” with cart contents and a line acknowledging them as a customer (“Good to see you back”). Goal: quick return to checkout.
• Email 2, 48 hours: reminder plus a genuinely useful detail — stock level if low, or a related item that pairs with the cart. Goal: recover without incentive.

Branch C: repeat customers (2+ orders)

• Email 1, 4 hours: short, warm, personal tone. Cart contents, one button. No urgency theatrics, no discount. Goal: nudge, protect margin, protect the relationship.

Channel is email for all branches to start. Once this works, SMS as a second touch for high-value carts is worth testing — see combining email and SMS for cart recovery. And if you’re unsure how many messages each branch can carry, how many abandoned cart emails an online store should send covers the trade-offs.

What this looks like in practice

Illustrative example: a store selling skincare at €55 average cart. A first-time visitor abandons a serum. Branch A’s first email tells her delivery is 2–3 days, returns are free for 30 days, and Klarna is available — the exact objections that made her hesitate. A customer on her third order abandons a restock of the same serum. Branch C sends one friendly reminder at hour four and nothing else; she completes the order at full price. Same trigger, two completely different conversations, no margin wasted. The numbers in this example are invented to show the mechanics, not promised results.

Four metrics that tell you if the split is working

• Revenue per recipient, per branch. The single best comparison number. Expect branch C to be highest per email sent.
• Recovered carts by segment. Watch whether branch A improves after you add trust content.
• Discount usage rate on recovered orders. This should fall versus your old generic flow.
• Repeat purchase rate. If regulars start abandoning more often after you remove their discount, revisit branch C’s tone — but usually this number holds.

Measure over at least 4–6 weeks; cart flows need volume before the branch numbers mean anything. If clicks look fine but orders don’t follow, why your abandoned cart emails get clicks but no orders walks through the usual causes.

Setting this up in Omnisend

In Omnisend, build this with one Cart Abandonment workflow and a conditional split on “number of orders placed” (0 / 1 / 2 or more), then attach the emails and delays above to each branch. The cart contents block pulls products in automatically, and the exit-on-purchase rule is standard. I resisted email marketing for years — I thought of it as the spam folder’s main supplier — until rising ad costs on my own stores forced me to look at behavioral flows. I tested Klaviyo and Omnisend and settled on Omnisend because I could build and adjust splits like this one myself, without a retainer. One honest caveat: the split is only as good as your data. If your platform integration doesn’t sync order history correctly, everyone lands in branch A, so verify a few known repeat customers before trusting the flow.

Your next step

Open your current cart flow report and check one thing: what share of recovered orders used a discount code, and how many of those buyers had ordered before. If that share is more than a token amount, build the three-branch split this week — branch C alone pays for the hour it takes.