Abandoned Checkout Recovery: The Full Guide for DTC Brands

A shopper fills a cart, types an email address and a phone number, gets to shipping or payment, and leaves. Most stores answer with the same reminder email for everyone, an hour later, maybe with a discount. A few of those shoppers come back. Most do not, because the reminder does not answer the question that stopped them.

The bigger cost is the shoppers the reminder should never have reached. The customer who has an open support ticket about a late order gets a cheerful nudge to finish checking out. The repeat buyer who completed the order on a laptop ten minutes later gets a "you left something behind" text on their phone. Each tool fires on the slice of data it can see, and the brand looks like it forgot the customer between messages.

This guide covers abandoned checkout recovery for Shopify and BigCommerce brands from start to finish: who to skip, when a reminder email is enough, when a trained person should text, which rules apply, and how to measure the revenue you actually recover.

What does the abandoned checkout journey look like for a DTC brand?

An abandoned cart and an abandoned checkout are different events. A cart is a list of products. A checkout is a shopper who has started to pay. Shopify records the second one as its own object, with the line items, prices, addresses, timestamps and a recovery URL the merchant can send so the buyer can pick up where they stopped³. Because the shopper has already typed contact details, a checkout is the event you can act on directly.

Abandoned Checkout emails are very profitable.

Why?

Because people forget to buy!

(we get distracted, forgetful, etc)

NOTE: This is different from Abandoned Cart - these customers have already started checking out.

Abandonment is common. Baymard Institute averages 50 studies and puts the documented cart abandonment rate at 70.22%¹. The stakes keep growing: the Census Bureau estimates US retail ecommerce sales of $340.2 billion in the second quarter of 2026, 17.1% of total retail sales on a seasonally adjusted basis².

Baymard also asks US shoppers why they abandoned an order in the past quarter. Leaving aside people who were just browsing, extra costs such as shipping, tax and fees lead at 40%. Slow delivery follows at 20%, then not trusting the site with card details at 19%, being asked to create an account at 18%, and a long or complicated checkout at 17%¹.

Two things follow from that list. First, some of these causes belong to the checkout itself. Baymard's usability testing finds that the average site has 32 unique improvements to make in its checkout flow, and that fixing them can lift conversion by 35%⁴. Recovery messages do not replace that work. Second, many of the remaining causes are questions: what will shipping cost, when will it arrive, what if it does not fit, is this site safe. A reminder that only says "you forgot something" answers none of them.

So the journey after a stall has three possible paths. Some shoppers should hear nothing yet. Some need a short reminder with the right facts in it. Some need a person who can answer a question. The rest of this guide is about choosing the path for each shopper, and the data each choice depends on.

How does Earth bring customer data into one profile?

You cannot pick the right path if every tool sees a different customer. The checkout lives in the store. The open ticket lives in the helpdesk. The subscription lives in the subscription app. The email address on the checkout may not match the one on last year's order. Earth, VYG's customer data platform, builds one living profile per customer from the systems you already run: orders, browsing, messages, tickets, subscriptions, reviews and loyalty. It matches the same shopper across devices and email addresses into one profile, and segments update themselves as the profile changes.

For abandoned checkouts, the profile answers the questions a flow should ask before it sends anything:

• Has the shopper ordered since? If the same person finished the purchase on another device or under another email, there is nothing to recover. Identity matching is what stops the "you left something behind" message to someone who did not.
• Is there an open support ticket? A shopper waiting on a refund or a late package does not want a checkout nudge. The ticket comes first.
• Is this a subscriber? An active subscriber adding a one-time item is a different conversation from a first-time visitor.
• What is the customer worth? Lifetime value and order history decide whether the next step is an email or a person.
• Have they opted out? Consent and opt-out history belong on the profile, not in one tool's settings.

Teams can also ask Earth these questions in plain English, including through Claude and ChatGPT over MCP. "Which checkouts from the last week came from customers with an open ticket?" is a list, not a project.

How does Jupiter automate the routine messages?

Jupiter is VYG's flow and automation layer, and it is in beta. Its flows read the whole Earth profile, so an abandoned checkout flow can branch on history (lifetime value, an open ticket, a refund) instead of the last click. One flow can mix email, SMS and human steps from Venus. The flow on the ecommerce industry page is the pattern most brands start with: wait while a Gorgias ticket is open, branch on lifetime value, send a reminder email through Klaviyo for most shoppers, and route the high-value checkouts to a person.

Situation on the profile	What the flow does	Why
Order placed since the checkout, on any device or email	Stops. No message.	There is nothing to recover, and a reminder reads as a mistake.
Open support ticket or refund in progress	Waits until the ticket is resolved, then re-checks.	The customer's problem comes before the sale.
New or low-value shopper, no question asked	Sends a reminder email with the recovery link, shipping cost and return terms.	A clear email is enough for a shopper who simply got distracted.
High lifetime value or a large checkout	Routes to a Venus agent who texts as the brand.	These shoppers are worth a conversation, and often have a question.
Active subscriber	Uses a subscriber message, or adds the item to the next shipment if the subscription app supports it.	The relationship already exists; a cold reminder ignores it.

When should the reminder go out?

Faster is not always better. A study in the Journal of Marketing ran field experiments where shoppers who abandoned carts were randomly assigned to receive a retargeting message or nothing. Shoppers who got an early message, 30 minutes to one hour after abandoning, were less likely to buy than the group that got nothing. A later message, one to three days after, had a positive effect, and a second experiment delivered by SMS replicated the pattern⁵. The authors also found the effects were larger for carts with more items and higher prices⁵.

That does not mean every brand should wait three days. It means timing is a variable to test, not a default to copy. Many shoppers do just forget, as the post above says, and a well-timed reminder catches them. A flow that sends within minutes to every shopper is the version most likely to annoy.

What should the reminder say?

Content matters as much as timing. A set of randomized field experiments presented at ICIS 2022 tested what cart recovery messages should highlight. Messages built on the shopper's cart history raised purchases, and the effect came from messages that highlighted return information rather than product reminders or shipping details. Net revenue for messages with return information was 49.7% larger than for messages with product information⁶.

Returns are a real worry for shoppers. In a National Retail Federation and Happy Returns report, retailers estimated that 16.9% of 2024 sales would be returned, and 67% of consumers said a negative return experience would discourage them from shopping with a retailer again⁷. If your return policy is good, say so in the reminder, in one plain sentence, with a link.

What about discounts?

The same Journal of Marketing research notes that early messages with a price discount or scarcity framing can draw more purchases, but warns that discounts used repeatedly can teach customers to wait for them⁵. A flow that reads the profile can hold the discount for shoppers who have never bought and keep it away from repeat customers who would have paid full price.

When should a real person step in with Venus?

Automation handles the shoppers who forgot. It does not handle the shoppers who stopped because of a question: will this match the table I already own, does the larger size run big, can it arrive by Friday, can I return it if it does not work. Those are the moments a template cannot read.

Every abandoned cart flow I've ever built is a monologue.

More emails, tighter timing, better subject lines, SMS, push. We keep adding layers, but the customer stopped listening after the second email.

Venus, from LiveRecover, puts a real, professionally trained person inside the brand's SMS flows. When a shopper stalls, pushes back or asks something off-script, a person reads the moment and replies. Agents are trained on the brand's products, policies and voice before they send a text, and they run alongside the brand's SMS marketing list. Every conversation and decision is written back to the customer profile in Earth, so the next flow knows what was said.

Text reaches nearly everyone who leaves a phone number at checkout. Pew Research Center reports that 91% of US adults own a smartphone, up from 35% in its first survey of smartphone ownership in 2011⁸.

Which checkouts should go to a person?

Not all of them. A person costs more per conversation than an email, so the flow should send the checkouts where a conversation is most likely to change the outcome:

• Checkouts above a value threshold you choose, or from customers with high lifetime value.
• Products that raise fit, color, sizing or compatibility questions.
• Shoppers who replied to an automated message with a question.
• First-time buyers who stalled at payment, where trust is the likely issue.

What does good look like?

Use clear benchmarks so the program can be judged. When a real person reaches out, a reply rate of 10-15% is good and 15-25% is excellent. For recovery rate, the share of conversations that end in an order, 8-12% is good and 12-20% is excellent. Brands typically see a 10-20X return on investment. Venus brands pay only on recovered sales, never per message, with no setup fees and no engineering needed.

How does FDE connect the systems you already run?

Every rule above depends on data the flow can actually read. FDE, VYG's forward deployed engineering service, is the part that makes that true. VYG engineers join your team in Slack, connect the systems your store runs on to Earth, build the first flows with you, handle web changes such as theme edits, landing pages and pop-ups, and set the team up on AI, for one monthly fee.

For an abandoned checkout program, the connections that matter most are these:

• Store and checkout: Shopify and BigCommerce connectors are live. WooCommerce and Checkout Champ are in beta.
• Email and SMS: Klaviyo, Infobip and Resend are live, so the reminder email can keep going out through the tool you already use.
• Helpdesk: Gorgias is live. Zendesk is in beta. This is the connection that lets a flow wait on an open ticket.
• Subscriptions: ReCharge, Skio, Smartrr, Stay AI and Ordergroove are live, so subscribers can get their own branch.
• Post-purchase and reviews: Loop Returns, AfterShip, Okendo and Yotpo are live, which puts return and delivery history on the profile.
• Anything else: a custom webhook and API connector is in beta, and CSV and SFTP imports and Zapier are built on request.

Part of the work is also removing what you no longer need. When one profile and one flow handle the checkout, overlapping apps that each sent their own reminder can be turned off, which is often how the duplicate messages stop.

What rules apply to texting shoppers who abandon checkout?

An abandoned checkout text is a marketing message, so consent rules apply. The details depend on the law and on how your messages are sent, and you should get your own legal advice. A few federal rules are worth knowing.

In 2024 the FCC adopted rules that let consumers revoke consent to robocalls and robotexts by any reasonable means, require callers and texters to honor opt-out requests within a reasonable time not to exceed 10 business days, and limit senders to a one-time text confirming the opt-out⁹. In practice that means a reply of "stop" or "unsubscribe" to any message has to work, and your flows need to read the opt-out wherever it was recorded.

The rules are still moving. In a 2026 order, reported in October, the FCC replaced the "revoke all" requirement, which had been on hold by waiver until January 31, 2027, with category-specific opt-outs for informational robocalls and robotexts. It also asked for comment on whether to shorten the deadline for honoring opt-outs from 10 business days to seven business days¹⁰. Check where these land before you build anything that depends on the timing.

Email reminders have their own rules: accurate sender details, an honest subject line, a working unsubscribe link and a physical address in marketing email. Quiet hours, state texting laws and carrier rules for business messaging add more. None of this is a reason to avoid recovery messages. It is a reason to run them from one place that knows who said yes and who said stop.

How do you get started?

Most brands can stand up a better abandoned checkout program in a few steps:

1. Connect the store, email tool and helpdesk. Shopify or BigCommerce, your email platform and Gorgias are the minimum for a flow that can skip the right people.
2. Write the skip rules first. Recent order on any device or email, open ticket, refund in progress, opted out. These protect the customer relationship.
3. Fix the reminder content. Total with shipping, delivery estimate, return policy and the recovery link. Keep discounts for shoppers who have never bought.
4. Set a value threshold for a person. Start with your top checkouts by cart value or lifetime value and route them to Venus.
5. Test timing. Compare an early send against a later one for the same audience rather than assuming faster is better.
6. Measure with a holdout. Keep a random slice of abandoned checkouts that gets no message.

How do you measure recovered revenue?

Last-click attribution overstates recovery. Some shoppers would have come back on their own, and a reminder that arrives just before they do gets the credit. The research cited above handled this with randomized control groups: some shoppers got a message, some got nothing, and the difference between the groups is the effect⁵⁶. You can do the same. Hold out a random slice of abandoned checkouts, for example one in ten, from all messages, compare the order rate in each group over the same window, and count only the difference as recovered revenue. Report it per path (email, person) so you know which one earns its cost.

Write every outcome back to the profile: who was skipped and why, who got which message, who replied, who ordered. Over a few months that history becomes the best guide to where your threshold for a person should sit.

If you want to see this running on your own store's data, book a demo and we will walk through your checkouts, your stack and the first flow with you.