WhatsApp Abandoned Cart Recovery Platforms for Shopify

How this list was selected

Use a common buyer scenario to compare what each included option publicly documents.

Selection method reviewed 2026-11-10.

Check which shoppers you could actually reach

Check the recipient's market first. Shopify says WhatsApp marketing in Shopify Messaging excludes recipients in the United States, while a US store can send to opted-in shoppers in supported countries. Meta classifies abandoned-cart reminders as marketing. Moving a marketing reminder to a different app or labeling it an order update is no basis for assuming it will go through. A US-only Shopify store should price its existing email or SMS route first. A US store with a sizable eligible international audience has a different decision. Shopify's regional eligibility rules and Meta's message categories.

Pull one typical month of abandoned cart or checkout events. Pick a unit you can reconcile, such as unique recoverable checkout IDs, and remove completed orders, duplicate events and any customer you could not identify. Next, check recipient markets and WhatsApp marketing permission for the remaining records. Keep a blank consent field as unknown. In a future trial, confirm that the store can show where permission came from and when it last changed. Shopify requires explicit WhatsApp marketing opt-in, and the platform policy calls for honoring requests to stop communications. Shopify's consent and sending considerations and WhatsApp Business Messaging Policy.

For example, say 1,200 unique checkouts stayed incomplete. In 900 of those checkout records, the store can identify the shopper. Of those records, 450 have a recipient market covered by the proposed setup. Of those 450, only 260 have a current WhatsApp opt-in that the proposed sender can see. Use 260 checkout records as the initial ceiling. Consolidate overlapping checkouts for the same shopper before sending and count deliveries separately. An order might arrive before the reminder, the approved template may not be available in the right language, or the platform can still limit delivery. Track each exclusion separately so the merchant can tell whether it needs a new tool, better consent capture or a different channel.

Ask who can access the WhatsApp Business Account and which number and account are covered in the proposal. Shopify Messaging documents connecting a WhatsApp Business Account through Meta's embedded signup and verifying eligibility during onboarding. Some merchant locations have extra conditions for sending outside their own country. Have a vendor confirm the target country, sender number, template and account before it prices the rollout. Shopify's store and account requirements.

Shortlist the platform that can own the next send

A merchant who needs one reminder should make the smallest working setup prove itself first. A merchant who needs coordinated email, WhatsApp and customer service has more to evaluate. Use the same sample cart in each demo. When the seller opens its flow, identify which store event starts it, which system decides who is eligible, who answers replies, and where you can see what happened.

How each provider earns a demo

Shopify Messaging

Start here when the gap is one automated reminder in a supported market. Shopify lists separate cart and checkout templates for WhatsApp. Its marketing setup verifies the connected WhatsApp Business Account. A separate inbox or reply workflow still needs an owner. Make the operator show the exact approved template and order stop before activating the route. Shopify's automation options and Shopify's account and market requirements.

Meta's WhatsApp app for Shopify

Consider it when you want a direct WhatsApp app for reminders and order updates. Meta lists prebuilt templates and automated abandoned-cart marketing in its own app. The app can generate separate Meta charges. Confirm that it has the event and support coverage you need before adding another sender to Shopify Messaging or an existing email program. Meta's Shopify app listing and billing notes.

Klaviyo

Compare Klaviyo if it already owns your email recovery flow. Its WhatsApp cart guide keeps the first email on the path for shoppers who cannot receive WhatsApp marketing, then rejoins the paths. A consent split does not settle what happens after a delivery failure. Test a valid WhatsApp recipient whose message then fails and check which system owns the next send. Klaviyo's recovery branch and order stop.

Wati

Compare Wati if you need to test an earlier cart event or configurable fallback. Wati documents separate cart and checkout recovery and a built-in rule with stop and fallback controls. The two documented workflows have different template and eligibility details. Ask which route is covered on the proposed plan and replay the cart-to-checkout transition on that route. Wati's recovery event conditions and Wati's built-in automation options.

Dondy

Shortlist Dondy when a Shopify-linked inbox should sit alongside automated recovery. The listing puts the integrated inbox and automated WhatsApp recovery on Power Automation and AI sales and support tools on Elite. The lower Pro card describes manual abandoned-cart recovery. Compare the plan that actually automates the send, then test who receives a policy exception. Dondy's Shopify plans.

Interakt

Consider Interakt if order notifications and a shared inbox belong in the same proposal. Its listing describes reminders with a checkout link, order notifications and FAQ auto-replies. The published Starter base leaves additional WhatsApp charges to confirm. Ask the vendor to quote the same event, markets and reply coverage as the other finalists. Interakt's app capabilities and plans.

Zobi

Consider Zobi when the buyer needs an answer before finishing the order. Zobi describes cart recovery and replies within one conversation, with WhatsApp among its channels. Confirm the exact sender, market and recovery event in the proposal. Zobi publishes this guide and offers shopper conversation software. This shortlist groups tools by job and does not rank provider performance. Start growing revenue.

Follow one shopper from cart to checkout to order

Ask every vendor to separate an add-to-cart event from an abandoned checkout. Shopify describes cart recovery as reaching shoppers who added a product without starting checkout and checkout recovery as reaching those who started checkout without ordering. The earlier event can add coverage only when the system can identify and contact that shopper. A dashboard with many added carts does not show how many of those shoppers are reachable. Shopify's recovery event definitions.

Wati's recovery documentation makes the difference concrete. Its add-to-cart route describes logged-in shoppers, current-session opt-ins and recognized returning visitors as ways to establish identity. Its checkout recovery instructions list specific inputs, including email, email marketing consent and a phone number. That is a documented condition of that Wati route. Keep WhatsApp permission as a separate requirement. Wati also distinguishes between fixed templates in its separate recovery documentation and configurable controls in its built-in cart rule. Before comparing providers, ask Wati which route is in the proposal and run the test on that route. Wati's identity, consent and checkout conditions and Wati's built-in rule configuration.

Use a test shopper in an eligible market who has valid WhatsApp permission. At 10:00, add a jacket in a size with known stock. At 10:10, start checkout. At 10:20, place an order before the next action is due. In a clean setup, starting checkout retires the earlier cart action if both routes are active, and placing the order stops any outstanding recovery for that order. Save the identifiers and timestamps in the merchant's source system and in each sending tool. A preview that shows a green check without an order event or a skipped action gives you less evidence than an actual test record. Wati documents options to stop when checkout starts or a purchase completes. Klaviyo documents an order filter in its WhatsApp cart flow. Wati's stop conditions and Klaviyo's order filter.

Run a second profile through the same path but let the checkout remain incomplete. Open the link in the message on a phone in the same market. Check item, variant, quantity, currency and any promised code. Then reopen it on a device where the shopper is not signed in. Find out whether the purchase path still works or whether the customer needs to identify themselves again. Ask what happens if a product becomes unavailable between the event and the send. Klaviyo's example can skip when an item is unpublished in its catalog, with the limitation that this is not a direct inventory quantity check. That is exactly the sort of difference a real cart reveals. Klaviyo's link and catalog behavior.

Verify the approved message and its checkout link

Test the actual template category, language and current approval before reviewing the copy. Shopify says its WhatsApp abandonment automation templates are submitted to Meta for approval. The WhatsApp policy calls for an approved message template when the business starts the conversation. A customer's inbound message can open a separate 24-hour customer service window. Treat the outbound cart reminder and the reply session as different events when you test eligibility, billing or what happens next. Shopify's template approval step and WhatsApp's template and reply rules.

Ask the vendor to demonstrate the version that would actually be sent in each launch language. Make sure the cart and checkout button refer to the shopper's original event and that claims about price, discounts or expiry are true for that cart. In Klaviyo's worked example, the link can come from the trigger event. Its instructions say to use expiry language only if the cart actually expires. Show a clean version with no first name and a mixed-item cart. If the experience depends on which product is displayed, set the order of items explicitly and confirm what the shopper sees. Klaviyo's template and checkout link example.

Test what happens when the language variant is unapproved, the template is paused or the store changes the offer. Some tools use one approved template with variables. Others may have different approved templates per message or market. Ask who can replace the template, how long that change can stay in review, and what is shown to operators while it is unavailable. Do not allow an unapproved template to go out to make up for a blocked send. A sensible launch can have fewer languages and a reliable way to pause than a full set of messages your team cannot support.

Decide what a failed WhatsApp send should do

Set one rule for an ineligible recipient and another for a confirmed delivery failure. A new fallback channel needs its own permission, a current cart, and a clear owner. It also needs to wait long enough for the primary status to become useful. If a system cannot tell you whether the WhatsApp action is pending or failed, switching to SMS immediately can create two reminders for the same shopper. Decide how long the team will wait and what status it needs before another tool can send.

Wati describes options to send a fallback for just the failed message or for all remaining messages, and describes stopping the rule after a reply as an option. Those are different choices with different consequences. If email already owns the second message, shifting the entire remaining sequence to SMS can duplicate that email even when the failed WhatsApp message is handled correctly. Klaviyo's documented first-message consent branch addresses entry to its email or WhatsApp path. Ask how the setup handles a later delivery failure after that decision. Wati's fallback and stop controls and Klaviyo's recovery branch.

Give the demo operator a late order as a negative control. Hold the fallback until the primary result is known. Then complete the test order while the fallback is still waiting and check that nothing sends. If the systems disagree about who placed the order, capture the matching profile, checkout and order references. Correct the integration or keep the old sender in charge. Canceling that one message by hand does not explain why it was still scheduled.

Give the person answering a reply control of the thread

Use two real categories of shopper question in the trial. A shopper wants to know whether the jacket fits over a midlayer. An international shopper asks whether a return would be free from their country. Give the operator the product fit notes and the return policy before the exercise. Deliberately leave one policy detail unresolved. Ask to see the answer, its source, the next owner if it cannot be answered and what happens to the existing follow-up.

The WhatsApp Business policy says a business can reply without a template within the 24-hour customer service window after a user message. If a business uses automation in that window, the policy calls for a clear, direct escalation path. Rehearse a reply that sits overnight and a separate case that reaches a person after the service window. Ask how the system tells the agent whether the window is still open and which approved action is available when it closes. A bot response, an inbox assignment and permission to send another reminder are three different records. WhatsApp's customer service and escalation rules.

Compare the cost and coverage of the reply job. Dondy lists AI sales and support tools on Elite, while the Power Automation listing names a Shopify-linked inbox. Interakt lists a shared team inbox on Starter and presents order messaging and auto-replies as app features. Zobi describes cart recovery and replies within ongoing shopper conversations across channels. Ask each finalist what a real operator sees, whether they can pause a reminder and who changes the approved product or policy information. If your existing team can handle the few replies that come in, an automated reminder and an agreed handoff may be enough. Dondy's tier details, Interakt's Shopify plans and Start growing revenue.

Get a quote for the same delivered messages

Compare a normal month and a busy one using the same qualifying audience, markets and sequence. Meta says it charges for messages delivered, with rates varying by recipient market and message category. An app or partner can add a plan charge, a markup, sending or automation fees, seats, AI work or support. The apparent price per conversation in a partner quote and the Meta price per delivered message are different units. Ask which parts appear on which invoice and whether a quoted rate already includes the network charge. WhatsApp Business Platform pricing.

In the earlier example, say 50 of the 260 checkout records become ineligible or are dropped during deduplication before the first action. That leaves 210 first attempts. Assume 180 first deliveries and, after rechecking eligibility, 60 second deliveries. The planning case then uses 240 delivered marketing messages. If 140 are UK deliveries and 100 are Germany deliveries, multiply each count by the rate quoted for that market and category. Record attempts separately, including any failed follow-ups the partner charges for. Use these example volumes to compare quotes and your own delivery logs to estimate actual traffic.

The plan names can conceal the feature you need. Dondy's Shopify listing shows manual recovery on its $6.99 monthly Pro card and lists automated WhatsApp recovery on Power Automation at $79.99 monthly. Meta's Shopify app is free to install, with potential separate Meta charges. Wati's pricing lists separate message billing and optional charges for areas including its Shopify integration, automation or AI. Interakt's listing shows a $21 monthly Starter plan and additional WhatsApp charges. Ask each provider for the exact plan and markets in your proposal. None of those base listings is a price for the same full recovery program. Dondy's published pricing, Meta's Shopify app billing notes, Wati's billing and add-ons and Interakt's Shopify pricing.

Put the result on two separate lines. The recurring line includes the correct plan, all quoted delivery and fallback charges, and the person or service taking replies. The setup line includes approved template work, integration and consent mapping, and a brief overlap if the old sender needs to stay available. Test the quote against a sample bill. Ask for a UK delivery, a Germany delivery, a failed action and a shopper reply. The evidence you want is the event record, the unit used for billing and where it lands on an invoice. A lower base is only meaningful once the two proposals cover the same work.

Choose a rollout you can reverse

Start with one supported market, one shopper cohort, and one cart or checkout event. Write down the old and new sender, the template, which other messages that profile may receive and who can stop them. Run the proposed tests while the established flow remains the fallback, with sends confined to the trial setup. Give the operator who owns the old queue permission to inspect and close any pending action before the new sender takes over.

Keep one simple launch record. Choose the system that holds the final consent state, the log that proves an order arrived and the operator who restores the old route. If the new platform cannot show the latest opt-out, a cross-device link opens the wrong variant, or a completed order still receives a reminder, stop the trial and identify the cause. Before rollback, stop the new sends and reconcile any new opt-outs, replies and orders. Turning the old flow back on without checking either queue can create the same problem you were trying to avoid.

Use one outcome measure and keep diagnostic counts beside it. The number of eligible carts, delivered reminders, useful replies, orders and faults can tell you where the flow needs work. Orders assigned to the tool are also useful for billing reconciliation. They cannot establish how many orders the tool added. Where the audience is large enough to support the comparison, keep your current recovery route as the baseline and assign some equally eligible shoppers to that route during the same period. Change only the part you are evaluating. Define the same order window, discount rules and contribution measure for both groups. A small or unbalanced test can still tell you if a route worked, while leaving business impact uncertain.

Keep Shopify Messaging or your existing email and SMS setup if the eligible WhatsApp audience is small or if the baseline handles the test cases at a price the store can justify. Move to a dedicated WhatsApp platform when a verified gap in reach, sequencing or reply ownership survives the baseline test. A good buying decision has a specific event, a supported audience, an owner and a verified stop rule. The worksheet lets the store compare all four before it pays for a second sender.

Download the buyer worksheet to keep the quote and test results together.

How the options were selected

The comparison uses public platform documentation and current provider listings for Shopify. It groups tools by the job they can be considered for and uses the example shopper and traffic figures to make the trial repeatable. Market access, account approval, plan details and actual performance depend on the merchant's configuration and contract.

Publisher disclosure

Zobi publishes this guide and is included for shopper conversations. The other providers are compared using their public documentation.