Abandoned Cart Discount: How Much Do You Actually Need?

You added a discount to your abandoned cart email because everyone said you should. You felt a bit sick about the margin. And then almost nobody used it.

So you assumed discounting does not work on your brand. It usually does. The offer was just too small to change anyone’s mind.

Quick answer: Most abandoned cart emails need a discount of around 10% before shoppers act on it. A 5% offer is often too small to shift a decision, so people ignore it and you lose the sale anyway. Test raising the discount on your final reminder, keep browse abandonment lower or discount-free, and make every code single use.

How much of a discount does an abandoned cart email actually need?

An abandoned cart flow, the automated email sequence that fires when someone adds to cart and leaves without buying, is usually the highest-intent flow in your account. These people picked the product, chose the size, and got as far as the checkout. They are not browsing. They are hesitating.

Which is exactly why a token discount does nothing. At 5% off, a $60 order saves someone three dollars. Nobody reopens an email and re-enters their card details for three dollars.

Five percent does not dangle the carrot hard enough.

I recently worked with a product brand whose abandoned cart flow had sat at 5% off for months. It sent plenty of emails. It was barely converting anyone. We lifted the offer to 10% off on the final reminder and the conversions moved straight away, on the same list, the same products and the same email copy.

Nothing about the flow was broken. The number on the button was just too small to be interesting.

Here is the honest maths most founders skip. If the discount is the thing that gets the order over the line, 10% of a sale you were going to lose is better than 100% of nothing. If the shopper was always going to buy, you have given away a small slice of one order. Over a month of recovered carts, that trade usually comes out well ahead.

Should you test raising your abandoned cart discount?

Yes, and it is one of the fastest tests you can run inside a flow that is already live.

Do it properly though. Do not change the offer, the subject line, the images and the send delay in the same week, because then you will never know which change did the work.

Here is what that actually looks like in your account:

• Change one thing. The discount amount on the final reminder. Nothing else.
• Leave the earlier emails alone. Email one should never carry a discount. It is a reminder, not a bribe. Give people the chance to buy at full price first.
• Run it for a full sales cycle. Two to four weeks depending on your volume, not four days.
• Compare placed order rate and revenue per recipient, not open rate. Opens tell you nothing about whether the offer worked.
• Watch your average order value. If it holds steady, the discount is buying you sales rather than discounting sales you already had.

If you are not sure whether the flow itself is set up correctly before you start playing with the offer, check the trigger first. I have written about that in more detail in why abandoned cart emails stop converting when the trigger is wrong. There is no point testing a bigger discount on a flow that is firing for the wrong people.

Does your browse abandonment flow need a discount at all?

This is where the advice flips.

A browse abandonment flow, the sequence that fires when someone views a product and leaves without adding it to cart, catches a much colder audience. They did not commit to anything. Plenty of them are window shopping, comparing, or landed from a link and bounced.

Browse abandonment usually sends to far more people than your abandoned cart flow. So every percentage point you hand out there gets multiplied across a much bigger group, and a decent chunk of them would have bought anyway.

Give your coldest audience your biggest discount and you will train your whole list to wait for one.

So the test in browse abandonment goes the other way. If it currently has 10% off, try 5%. If it currently has no discount, try a small one and see whether the extra orders cover the margin. Both are worth knowing.

Better still, do not give it to everybody at once. If your flow already has an A/B split, send the discount down one path only and leave the other path selling on the product, the reviews and the reason to buy. That way you learn whether the discount is doing anything, and you only pay for it on half your traffic while you find out. There is more on setting these up cleanly in my guide to A/B testing Klaviyo emails properly.

Same thinking applies across every automation you have running. If you have not mapped out what each one should be doing yet, start with the 6 core Klaviyo flows every ecommerce store needs and work out which ones genuinely need an offer attached and which ones are fine selling on their own merit.

Why are shoppers using a discount code you never sent them?

Here is the leak almost nobody checks, and it is quietly costing you subscribers.

Your welcome pop-up offers 10% off in exchange for an email address. Fair trade. You get a subscriber, they get a discount, and your list grows.

Except your affiliate codes, your influencer codes and your old campaign codes are all sitting on the open internet. Coupon aggregator sites scrape them, publish them, and rank for them. So when a shopper hits your checkout and sees an empty discount box, they open a new tab, search your brand name plus “discount code”, and paste in the first one that works.

They get the discount. You get nothing. No email address, no subscriber, no welcome flow, no future campaigns to that person. You paid the margin and got none of the asset.

Every publicly indexed code is a subscriber you paid for and did not receive.

What to do about it:

• Rotate affiliate and influencer codes every one to two months. The scraped versions expire and the search results go stale. It is a small admin job with a real return.
• Make welcome pop-up codes single use and unique to the subscriber. A shared code like SAVE10 will be public within weeks. A unique code cannot be reused or passed around.
• Set an expiry. Seven to fourteen days on a welcome code adds genuine urgency and keeps old codes from circulating forever.
• Use auto-applied discount links in your emails so the offer lands in the cart when they click, instead of handing them a code they can screenshot and share.
• Search your own brand name plus “discount code” once a month. Whatever comes up on page one is what your customers are using. If you have never done this, do it today. It is confronting.

If you are looking at your own account right now wondering how many of these leaks you have, that is exactly what a Mini Klaviyo Audit is for. It is $69 and you get a 15-minute personalised video walking through where your account is losing money. No call, no pitch, just the list of what to fix.

How do you stop the same customer claiming your welcome discount over and over?

The other half of the leak is repeat claiming. One shopper works out that a new email address means a fresh welcome discount, and suddenly they are ordering every week at 10% off under a slightly different Gmail.

It is annoying. It is also more expensive than the discount itself, because of what it does to your list.

Every throwaway address sits in your account as a profile you are paying for. Half of them never open anything. Some of them bounce. Your failed delivery rate creeps up, your engagement rate drifts down, and the inbox providers quietly start treating your sending domain with more suspicion.

So you are paying three times. Once in margin, once in subscription cost, and once in deliverability.

The fixes are boring and they work:

• Unique, single-use codes on the pop-up so one address equals one discount.
• A minimum spend on the welcome offer, set just above your average order value.
• Limit the code to first-time customers in your ecommerce platform settings where that option exists.
• Run a sunset flow so unengaged and fake profiles are removed rather than sitting there costing you money and dragging your sender reputation down.
• Check that your sign-up forms all collect a first name. Footer sign-up boxes are often email only, which is why you end up with a list full of nameless profiles you cannot personalise to.

And a small operational note that saves a surprising number of support emails. Make sure the welcome email carrying the code sends immediately, not after a delay. A large share of “I never got my discount code” messages are simply people who checked their inbox before your flow sent.

What should your discount ladder look like across your flows?

Discounts should get bigger as intent gets higher and as the chance of losing the sale gets higher. Most brands have it backwards, handing the same offer to everyone regardless of where they are.

A sensible starting structure for a product brand:

1. Welcome flow: your headline offer, usually 10% off, single use, with an expiry. This is the one you advertise, because it buys you a subscriber.
2. Browse abandonment: no discount, or a small one on a split path only. These people have shown the least intent.
3. Abandoned cart: no discount on the first reminder. Discount on the final reminder, usually around 10%, because this is a sale you are about to lose entirely.
4. Post-purchase and retention: value before offers. Care instructions, styling, how to get more out of what they bought. Save the discount for the win-back, when they have genuinely gone quiet.
5. Win-back: your most generous offer, aimed at customers who have already proven they will spend with you.

Notice what that does. The people most likely to buy anyway get the smallest nudge. The people you are about to lose get the strongest one. That is the whole idea.

Key takeaways

• A 5% abandoned cart discount is usually too small to change a decision. Around 10% on the final reminder is where most product brands see movement.
• Test the offer amount on its own, over a full sales cycle, and judge it on placed order rate and revenue per recipient rather than opens.
• Browse abandonment reaches a colder, larger audience, so it should carry a smaller discount or none at all, ideally tested on one split path first.
• Publicly scraped affiliate and campaign codes let shoppers skip your pop-up entirely, so you pay the margin and never gain the subscriber.
• Rotate affiliate codes, make welcome codes unique and single use, add expiries, and search your own brand plus “discount code” monthly.
• Repeat discount claiming with throwaway email addresses costs you margin, subscription fees and deliverability all at once.