# Should Abandoned Cart Texts Include a Discount?

> Many unfinished checkouts stall over questions, not price. Here is when an abandoned cart discount actually helps, and how to keep any incentive small.

Canonical: https://www.vyg.ai/blog/abandoned-cart-discount
Last updated: 2026-11-19

**Short answer:** Usually not by default. Shoppers who leave a checkout most often cite extra costs like shipping, slow delivery, trust worries or a long checkout, and a clear answer fixes most of those. Save a small, single-use code for shoppers genuinely comparing prices, never stack codes, and never invent a former price.

**If this sounds like your store:** every cart recovery text you send carries a discount code, your margins keep shrinking, and you suspect some shoppers have learned to wait for the code before they buy. You want to know whether the discount is doing the work, or quietly paying for sales you would have won anyway.

Attaching a discount to every cart reminder feels like the safe move. A code is easy to set up, easy to measure, and it gives a stalled shopper an obvious reason to come back. It also feels like doing something. So most stores reach for it by default: a checkout goes quiet, and out goes a text promising a percentage off.

The trouble is that a blanket code answers a question many shoppers never asked. It teaches your best customers to wait for the offer, it eats the margin on orders that were already close to the line, and it does nothing for the shopper who left for a reason that has nothing to do with price. The reflex is understandable, but it is worth questioning. Before you decide what an abandoned cart discount should do in your next recovery text, it helps to look at why checkouts actually stall.

## Why do shoppers leave a cart, if not for price?

Carts going unfinished is the normal state of online retail, not an emergency. Across 50 studies, the average documented cart abandonment rate is 70.22%. 1 Treating every unfinished cart as lost revenue to be bought back with a code overstates both the problem and the fix. Many of those shoppers were just browsing, so the real question is narrower: of the ones you can still win, how many actually need a lower price?

When shoppers who are past the just-browsing stage are asked why they left, the answers are mostly about friction and trust, not the sticker price. Extra costs that show up at checkout lead the list at 40%, followed by delivery that is too slow at 20%, not trusting the site with card details at 19%, being forced to create an account at 18%, and a checkout that is too long or complicated at 17%. 1

Cost and friction at checkout, not the product price, stop most shoppers. Source: Baymard Institute 1.

Only the first of those, extra costs, is really about money, and even that is usually a shipping or fee problem rather than the price of the product itself. A flat code does nothing for a slow delivery estimate, a clunky checkout, or a wobble over trust. The same is true of the shopper worried about returns or payment: they are weighing risk, not cost. A clear answer solves those (more in [why shoppers leave a cart](https://www.vyg.ai/blog/reasons-for-cart-abandonment)). That is the first reason most recoveries never need a code: the thing in the way is a question, not the price.

## When does an abandoned cart discount help?

Sometimes price genuinely is the issue, and a discount is the right tool. The useful part is how little it takes. In a March 2026 Bizrate Insights survey of 1,010 U.S. adults, 43% said a discount in the 10 to 20% range is the minimum needed to move them from the fence to a finished order, and 34% named the next band up. 2 Deep cuts are rarely what tips a decision.

When a discount tips an undecided shopper, a shallow one usually does it. Source: Bizrate Insights 2.

So when you do reach for a code, keep it shallow and keep it rare. A modest, single-use code aimed at the shopper who is genuinely comparing you with a cheaper rival will protect your margin far better than a standing offer that everyone learns to expect. A discount you send to everyone is not really a recovery tool; it is a price cut with extra steps. Reserve the incentive for the cases where price is the real objection, and answer everything else with information.

Friction can even look like a price problem. The same survey found that a $10 return fee on a $50 purchase is enough to drive 72% of shoppers to walk away from the order. 2 That shopper does not need a discount; they need to hear your return policy. Fix the fee or explain it, and the objection often disappears. Matching the response to the real reason is the whole game.

## When should you send a discount code?

You do not need a complicated policy. You need a quick read on why this particular shopper stopped, and a habit of answering first and discounting second. Rules can sort the easy cases, and the gray ones need a little judgment. The table below is a starting point for the most common situations.

| What stopped the shopper | Better response |
|---|---|
| Shipping or fees surprised them at checkout | Show the free-shipping threshold or offer flat-rate shipping; a code is rarely the fix |
| They did not trust the site or hit a payment error | Reassure them and clear the error; a discount does not build trust |
| They are comparing you with a cheaper rival | A small, single-use code can close the gap |
| A first-time buyer is hesitating on a higher-priced item | A modest first-order code can tip the decision |
| They got distracted and are not ready yet | Send a reminder and a question; hold the code |
| Something went wrong (late, damaged, or wrong item) | A goodwill code is good service, not acquisition |

The pattern holds across all of them: when the blocker is a question, answer it; when it is genuinely price, a small code can close the gap; and when something went wrong on your side, a goodwill gesture is about service, not acquisition. Written down, the rule is almost boring, which is the point.

## Make the code safe when you send one

A discount that leaks or misleads costs more than it ever recovers. A few guardrails keep incentives under control:

- **Make every code single-use and short-lived.** A unique code that expires in a day or two recovers the sale without turning into a permanent price cut.
- **Do not let codes stack.** One recovery offer per order stops a small nudge from quietly compounding into a much deeper cut.
- **Be honest about the former price.** If you present a code as a saving off a former price, that former price has to be a bona fide price the item was openly offered at, not a fictitious higher number set up to make the deal look bigger. 3
- **Tie each code to a reason.** A code that answers a real objection protects your margin; a code sprayed at everyone just trains shoppers to wait for the next one.

## Where a real person fits

Almost everything above comes down to one skill: reading why a shopper stopped and responding to that, rather than reaching for the same code every time. An automated reminder cannot do it, because it can only send what it was told to send in advance. A person can. It is the difference between a message and a conversation.

VYG Conversations puts a real, trained person inside your cart and checkout texts. When a shopper stalls or pushes back, a Conversations agent reads the moment and answers the real objection, whether that is shipping, sizing, a trust worry, or a genuine question about price. The agent is briefed on your products, policies, and voice first, so the reply sounds like you and never like a script. Because a person can answer the actual question, the discount can stay in reserve for the shoppers who truly need it. For wording ideas, see [how to write cart recovery texts that get replies](https://www.vyg.ai/blog/how-to-write-abandoned-cart-texts-that-get-replies).

The numbers worth watching here are conversation numbers. When a real person reaches out by text, a reply rate of 10-15% is a good result and 15-25% is excellent; of the conversations that follow, 8-12% ending in an order is good and 12-20% is excellent. Brands typically see a 10-20X return on those conversations. Pricing depends on your setup, so see the [pricing](https://www.vyg.ai/pricing) page.

See what that looks like on your own store: watch a real person turn stalled checkouts into orders in a short [Conversations](https://www.vyg.ai/conversations/demo) demo.

## Frequently asked questions

### Should every abandoned cart text include a discount?

No. A code sent to everyone teaches shoppers to wait for it and cuts margin on orders that would have closed anyway. Answer the shopper's question first and keep the code for real price objections.

### How big should an abandoned cart discount be?

Small. In a 2026 Bizrate Insights survey, the largest group of shoppers said a modest discount was enough to move them from on the fence to buying. Deep cuts are rarely needed.

### How do I stop discount codes from leaking?

Make each code single-use, set it to expire in a day or two, and do not let it stack with other offers.

### Can I show a discount off a higher original price?

Only if that original price is one the item was genuinely offered at. The FTC's Guides Against Deceptive Pricing treat a made-up former price as deceptive.

## Sources

1. [Cart Abandonment Rate: 50 cart abandonment statistics](https://baymard.com/lists/cart-abandonment-rate), Baymard Institute, 2024.
2. [Free shipping beats lowest price in online purchase decisions](https://bizrateinsights.com/free-shipping-beats-lowest-price-as-the-1-driver-of-online-purchase-decisions/), Bizrate Insights, 2026.
3. [16 CFR Part 233: Guides Against Deceptive Pricing](https://www.ecfr.gov/current/title-16/chapter-I/subchapter-B/part-233), Electronic Code of Federal Regulations, 2024.

## Related

- [How Much Does SMS Marketing Cost? A Pricing Guide](https://www.vyg.ai/blog/how-much-does-sms-marketing-cost)
- [RCS Abandoned Cart Messages: Rich Cards, Real Replies](https://www.vyg.ai/blog/rcs-abandoned-cart)
- [When to Send Abandoned Cart SMS: Timing That Gets Replies](https://www.vyg.ai/blog/best-time-to-send-abandoned-cart-sms)
- [All posts](https://www.vyg.ai/blog)
