---
title: "Abandoned baskets: diagnosis and recovery · Lyntev"
description: "Investigate shop abandonment without confusing interest, payment errors and bots. Includes diagnosis, tests and criteria for recovery messages."
url: "https://lyntev.com/en/guides/diagnose-recover-abandoned-baskets"
language: "en-GB"
date_published: "2026-11-21T17:20:13.047Z"
date_modified: "2026-11-21T17:20:13.047Z"
---

1. [Home](<https://lyntev.com/en>)
2. [Practical guides](<https://lyntev.com/en/guides>)
3. Abandoned baskets: find the cause before sending reminders

# Abandoned baskets: find the cause before sending reminders

An incomplete basket may mean comparison, missing information, a payment error or automated activity. Start by defining your platform’s metric and trying checkout. Fix obstacles before sending reminders. Every message needs a recipient and a purpose for which you can legitimately use their contact details. Useful recovery is a completed order that is followed through, rather than merely an opened email or a returning visit.

By [Lyntev](<https://lyntev.com/en/company>). Updated on 20 November 2026.
About 6 minutes to read.

[Start reading](<https://lyntev.com/en/guides/diagnose-recover-abandoned-baskets#leitura>)[Download worksheet](<https://lyntev.com/recursos/en/recuperar-carrinhos-abandonados.txt>)

**From data to the next decision**Choose the question, define the calculation and act on the result.

Translation published on 21 November 2026. [Original Portuguese guide](<https://lyntev.com/guias/recuperar-carrinhos-abandonados>); source revision 20 November 2026.

By the end, you can:

- Define what counts as abandonment on your platform
- Investigate delivery and payment friction
- Assess recovery without confusing clicks with orders

In this guide

1. [Know what the platform is counting](<https://lyntev.com/en/guides/diagnose-recover-abandoned-baskets#definicao>)
2. [Group signals by a testable hypothesis](<https://lyntev.com/en/guides/diagnose-recover-abandoned-baskets#causas>)
3. [Protect people who may already have paid](<https://lyntev.com/en/guides/diagnose-recover-abandoned-baskets#pagamento>)
4. [Check the basis and context before contacting anyone](<https://lyntev.com/en/guides/diagnose-recover-abandoned-baskets#mensagens>)
5. [Fictional scenario: a fabric shop](<https://lyntev.com/en/guides/diagnose-recover-abandoned-baskets#cenario>)
6. [Measure the full process and unwanted effects](<https://lyntev.com/en/guides/diagnose-recover-abandoned-baskets#avaliar>)
7. [Create a diagnosis worksheet for each hypothesis](<https://lyntev.com/en/guides/diagnose-recover-abandoned-baskets#ficha>)
8. [Sources and criteria](<https://lyntev.com/en/guides/diagnose-recover-abandoned-baskets#fontes>)

## Know what the platform is counting

Distinguish a product added to a basket, checkout started, contact details entered and an order created. Tools may use different definitions and time windows. The Shopify documentation consulted presents recovery experiences with different settings and states that not all automated activity is filtered. Do not copy another platform’s definition into your reports without checking what the events represent.

Also record observation limits. Rejection of optional statistics, switching devices and collection failures may fragment the journey. Use permitted measurement to identify questions and tests, rather than claim to know every person’s motive. No purchase in a session does not prove the buyer rejected the product or that a reminder campaign could resolve the situation.

## Group signals by a testable hypothesis

Start with conditions appearing late: delivery cost, timescale, availability, mandatory account or payment method. Try the journey with baskets and destinations the shop serves. If a problem always occurs in a specific case, there is a demonstrable correction. If you merely observe fewer purchases, keep several hypotheses open and seek more information before changing the offer.

Talk to support about questions and incidents without copying personal data into a marketing spreadsheet. Separate technical failure, a commercial question, unavailable products and normal comparison. A recovery message does not replace fixing a failed button or incorrect total. It may even worsen the experience by inviting another payment attempt while one remains pending.

Group signals by a testable hypothesis

| Signal | Hypothesis | Check |
| --- | --- | --- |
| Exit when delivery is shown | Unexpected cost or timescale | Review earlier information and test destinations |
| Repeated payment attempts | Failure, slow confirmation or decline | Check statuses with the provider |
| Variant disappears in the basket | Selection or availability error | Repeat the selection and go back |
| Many events without plausible behaviour | Automated activity or incorrect collection | Check reports and technical signals |
| Later return without purchase | Comparison or insufficient information | Observe questions and tasks without inventing motives |

## Protect people who may already have paid

Before treating a checkout as a sales opportunity, check whether an operation is pending or captured. Some payment methods depend on later confirmation. Buyers may have closed the page and received information from their bank before the shop’s visible summary updates. Support must be able to link order and operation to avoid encouraging an improper second payment.

Establish an exclusion rule for orders being reconciled. If there is an incident, the message should support that specific case and match what the team knows, rather than pretend to be an automatic promotion. Limit access to necessary records. Do not use screenshots with full card numbers, credentials or other sensitive data to discuss the problem with a wider team.

## Check the basis and context before contacting anyone

Entering an email during purchase should not automatically be treated as permission for any campaign. CNPD distinguishes the direct electronic marketing regime and explains relevant conditions for individuals. Applying it to your case requires checking the purpose, existing relationship and applicable rules. Do not turn a technical option to contact everyone into a legal decision.

If there is an authorised, suitable recovery journey, define the content, frequency and stopping method. Identify the shop and link to a page with truthful conditions. Do not say stock is reserved unless that reservation exists. Do not invent a discount deadline. Separate requested support from promotional communications and confirm the treatment with the person responsible for business privacy.

## Fictional scenario: a fabric shop

A fictional shop observes several incomplete checkouts for fabric rolls. The team considers an automatic discount but first tests a basket with different lengths. It finds the product unit is unclear in the summary. Some people may not understand whether quantity means metres or pieces, so the team corrects that information before encouraging a return.

The shop also identifies a destination where delivery charges appear only at the final step and makes that condition available earlier. During evaluation, it records eligible sessions, checkouts, operations and orders received according to platform definitions. It does not attribute every subsequent change to the corrections, because the catalogue and demand also vary. The scenario is fictional and provides no expected recovery rate.

## Measure the full process and unwanted effects

When testing a recovery action, track messages actually delivered, return visits and confirmed orders, using relevant definitions and permissions. Do not count all subsequent purchases as caused by the reminder. The person could return through another channel. With low volume, observe functionality and feedback before seeking statistical differences the data cannot support.

Include complaints, requests to stop, support contacts and time spent resolving questions. An action producing many clicks while confusing buyers has not been properly resolved. Avoid successive campaigns to compensate for incomplete commercial information. The aim is to support a decision that already made sense to the customer, preserving trust and the team’s capacity to follow every order.

- The abandonment metric has a documented definition.
- Tests cover the variant, delivery, error and pending payment.
- Contact is not enabled merely because the tool allows it.
- Analysis distinguishes a confirmed order, a return visit and a simple interaction.

## Create a diagnosis worksheet for each hypothesis

Choose one hypothesis at a time and write how you will check it. This could be a functional test, information review or task observation. Record the change, date and other events that might affect results. This lets you learn from the test and avoids changing price, delivery, form and messages simultaneously without understanding what solved the difficulty.

Before expanding automation, ask the team to simulate someone who has already paid, someone who asked not to receive messages and someone who chose a variant now unavailable. Behaviour in these cases shows whether the rules are clear enough. If you cannot confirm the situation, keep sending paused and investigate the process. Technical ability to send does not replace appropriate contact.

### Worksheet for your project

```
Abandonment diagnosis
Platform definition and time window:
Observed signal and data limitations:
Main hypothesis:
Basket and destination used in the test:
Payment statuses checked:
Proposed correction and responsible person:
Permitted processing of contact details:
Content, frequency and stopping rule:
Outcome metric and side effects:
Date, other changes and possible conclusion:
```

## Sources and criteria

Content prepared with AI assistance and research using technical sources, with checks of the relevant examples and calculations. The references support the guide’s decisions; examples identified as fictitious let you try out the method.

- [Shopify — recuperação e definições de checkout abandonado; confirmar a experiência em utilização](<https://help.shopify.com/en/manual/promoting-marketing/create-marketing/abandoned-checkouts/>)
- [CNPD — Diretriz/2022/1 sobre comunicações eletrónicas de marketing direto](<https://www.cnpd.pt/umbraco/surface/cnpdDecision/download/121958>)

## Continue to your next decision.

- [Prepare payments and order statuses](<https://lyntev.com/en/guides/prepare-mb-way-multibanco-online-shop>)
- [Diagnose visits without enquiries](<https://lyntev.com/en/research/website-visits-without-enquiries>)
- [Explore online shop options](<https://lyntev.com/en/website-design/online-shops>)

This guide is part of [Sell through an online shop.](<https://lyntev.com/en/guides#lojas-online>)

- [How to show delivery charges and times without surprising customers](<https://lyntev.com/en/guides/delivery-charges-times-online-shop>)

[Explore the guide library](<https://lyntev.com/en/guides>)

## Would you like to apply this to your business?

Send us your worksheet or explain what you need to improve. We help you choose the next step for your project.

- [Lyntev assistant Immediate answers about services, prices and timescales, with a cost estimate. Read the assistant guide](<https://lyntev.com/en/help/ai-assistant>)
- [WhatsApp +351 928 412 891. The message is only sent when you confirm it in WhatsApp. Write on WhatsApp](<https://wa.me/351928412891?text=Hello!%20I%20came%20from%20the%20Lyntev%20website%20and%20would%20like%20to%20discuss%20a%20project.>)
- [Form Describe what you need and receive the reply by email. Complete the form](<https://lyntev.com/en/contact>)
- [Email contacto@lyntev.com Send an email](<mailto:contacto@lyntev.com?subject=Request%20for%20information>)

![Wellinton Oliveira](<https://lyntev.com/media/autores/wellinton-oliveira.webp>)

Written by

[Wellinton Oliveira](<https://lyntev.com/en/company>)

CEO of Lyntev

Technology with context, original analysis and sources you can consult.
