A Klaviyo flow not sending does not always mean email delivery is broken. The person may never have qualified, may be waiting for a scheduled step, may need manual review, or may have been intentionally excluded. Start by locating one profile’s journey instead of changing every message.
Quick answer
The direct solution is to trace entry, routing, message status, eligibility, and the recorded outcome in that order. Restore the specific broken condition, then test with an authorized profile. Do not remove suppression or protective filters just to make the send count increase.
Key takeaways
- Follow one profile through entry, routing, timing, eligibility, and message outcome.
- Read the recorded waiting, review, or skipped reason instead of assuming delivery failed.
- Keep suppression and protective filters intact while testing an authorized profile.
Documentation reviewed October 9, 2026. Scenarios are illustrative, not hands-on test results. Interface labels and channel availability can vary.
Choose one profile and one expected message
Begin with a precise statement: this profile performed this action at this time and should have reached this message under these rules. A complaint that “the welcome series is broken” is too broad to test. It could describe a missing signup event, an intentional delay, or an email that was sent but not noticed.
Use an approved test profile or a real incident that you are authorized to inspect. Do not add customers to lists, change their permissions, or trigger marketing messages merely to gather evidence. Record only the profile information necessary for troubleshooting and keep it out of public screenshots.
Write down the flow name, intended audience, trigger, expected route, and message name. Save the current configuration before editing. If several people administer the account, nominate one incident owner so another administrator does not change a split while you are testing its original behavior.
For teams deciding between systems, our Klaviyo versus Mailchimp comparison covers selection context. This guide is narrower: it diagnoses an existing Klaviyo journey without assuming that replacing the platform is necessary.
Classify the outcome before attempting a fix
Distinguish not entered, waiting, needing review, skipped, and sent. These are different investigation paths. Do not use an open-rate report as proof that a message was not sent. Likewise, a flow-level activity total cannot establish what happened to a particular person at a particular step.
| Evidence | Question to answer | Avoid |
|---|---|---|
| No journey identified | Did the expected trigger occur? | Editing message copy first |
| Waiting for a step | Is the scheduled time appropriate? | Removing every delay |
| Needs review | Was manual approval intended? | Assuming automatic delivery |
| Skipped | What recorded reason applies? | Disabling all protections |
| Sent but not received | What delivery evidence exists? | Re-entering the entire flow |
Use this table to choose your next question, not to infer a cause without evidence. A waiting person might be correctly scheduled. A skipped person might have completed a purchase and no longer need a reminder. Those outcomes can indicate that the flow is protecting the reader from an inappropriate message.
Check whether the person entered the intended flow
Compare the actual source event or membership change with the configured trigger. Similar names are not sufficient: two integrations can use different events to represent related business actions. Record which integration produced the event and whether its relevant properties match the intended rule.
Klaviyo’s trigger and filter documentation distinguishes entry-time trigger filters from profile filters, which are checked at entry and before flow actions. A person can fail eligibility for an action and continue to a later step where eligibility is checked again. This matters when interpreting an apparently missing message.
Do not remove an entry rule because your test profile happens not to satisfy it. First ask whether the rule represents the intended audience. If the rule is correct, change the test fixture or recognize that the person should not enter. If it is wrong, document the business decision that justifies the correction.
For a list-based welcome journey, trace the form’s destination and the actual membership change rather than repeatedly submitting the same address and assuming each submission creates a fresh journey. For a metric-based journey, locate the relevant recorded event. These checks keep source-data problems separate from message configuration.
Inspect the exact message’s status and timing
According to Klaviyo’s getting-started guide, draft messages do not schedule recipients, manual messages require review rather than automatic sending, and live messages send automatically to qualifying recipients. Check the status of the specific action, not just the flow’s name or overall appearance.
The flow-enablement instructions note that recipients already awaiting manual review still require review after a message is changed to live. Unconfigured actions can also remain draft after a bulk update. Avoid promising that a status change will automatically deliver every historical message.
Map the delays and branches between entry and the expected message. Record the person’s expected arrival time and the relevant account or configured scheduling timezone. For a business serving North America, the UK, and Australia, do not mistake a local clock comparison for evidence of failure.
If a delay was intentionally added to wait for a purchase or another event, shortening it changes the marketing strategy as well as the test. Use a controlled test route where appropriate rather than accelerating the live audience. Review the actual waiting state before deciding whether any scheduling adjustment is needed.
Read the skipped reason instead of guessing
Open the affected flow message, choose View details, and inspect Recipient activity. Select Skipped and use the reason filter to locate the relevant profile. Record the message and reason together. These navigation steps come from Klaviyo’s troubleshooting instructions linked below; if your account’s labels differ, use the equivalent message-level recipient view rather than a campaign report.
Klaviyo directs administrators to the message’s recipient activity and skipped reasons in its skipped-profile troubleshooting guide. Documented causes include Smart Sending, failed filters, suppression, invalid addresses, and message-specific configuration problems. The reason attached to your profile should guide the repair.
Classify that reason into one of three groups: an intended audience protection, a configuration mistake, or invalid source data. The same visible outcome can demand different action. A customer who has opted out should remain excluded. A malformed address requires correction through an approved process, not a switch that forces sending.
For a protective rule, ask whether the business expectation was wrong. A recent campaign may make another promotional message unnecessary, so a skip can be reasonable. When evidence confirms a configuration mistake, review the audience impact and change only the affected setting. Invalid input also requires correction in the upstream collection process.
Never treat email and SMS eligibility as interchangeable. Review the relevant channel, country, approved consent records, and current compliance requirements with the responsible owner. This article does not establish permission to contact anyone. A successful technical test should use a profile authorized for the channel being tested.
Review splits and content without bypassing the journey
Draw the actual route the profile took. For each branch, write the condition and the evidence at the time of evaluation. Do not assume a profile’s current properties perfectly represent what was true earlier. If historical evidence is unavailable, state that limitation and reproduce the case with a controlled fixture.
Inspect the intended message’s sender details, links, dynamic content, and required configuration. An editor preview can help catch missing text or an irrelevant product reference, but it does not establish that a person entered the flow or was eligible to send. Keep content preview and journey validation as separate checks.
Our email sequence guide is useful if the investigation reveals a confusing journey. Simplifying a sequence can improve maintainability, but do not redesign several active flows while diagnosing one missing message. Resolve the incident first and plan wider editorial work separately.
If the record indicates a send rather than a skip, move to delivery investigation. Our email deliverability guide provides that broader context. Re-entering the person into the sequence can create additional messages without solving the original delivery problem.
A fictional case: a live flow with a manual first message
Imagine a retailer expecting a new subscriber to receive a welcome discount immediately. A later educational message is live, but the first message is manual. The subscriber appears in review rather than receiving the expected email. Initially, the team suspects the signup form because the inbox remains empty.
The diagnostic record establishes that the subscriber joined the intended audience and reached the correct step. That evidence rules out the form as the immediate cause. The issue is the difference between the team’s expectation of automatic delivery and the action’s actual review setting.
The administrator reviews why manual status was chosen. Perhaps the discount content was awaiting approval; changing it to live without that approval would be inappropriate. Once approved, the team decides how to handle existing review items and how future eligible subscribers should progress. It does not assume one status change resolves both populations.
A fresh, authorized test then demonstrates the intended future journey. A separate review documents what happened to the original profile. This distinction avoids claiming a repair based on a new subscriber while leaving the historical case unresolved. The scenario illustrates the method, not a guarantee about your account.
Run a small acceptance test and preserve an exclusion test
Create a short test plan with a qualifying profile, a profile that should be excluded, and a profile that follows a different meaningful branch. Define the expected outcome for each before executing it. A flow that sends to everyone may pass a superficial send test while failing the business requirement.
Check the recorded journey, scheduled time, recipient outcome, and message content. For the excluded fixture, confirm that the protection remains intact. Use manual review or another approved containment method when testing could otherwise contact a real audience. Do not leave temporary test settings active after the exercise.
Save the before-and-after configuration and a brief incident note. Include the original symptom, profile evidence, root cause, authorized change, and acceptance result. Remove unnecessary personal information from the note. If the fix affects other flows using the same source data, review those deliberately rather than applying an indiscriminate account-wide change.
Use our email marketing metrics guide when returning to ongoing reporting. Track intentional exclusions separately from configuration failures so a low send count is not automatically interpreted as a broken automation.
Answers to common troubleshooting questions
Should I switch everything to live?
No. Check whether content and audience rules are approved, then change only the intended actions. Review pending historical cases separately. A bulk status change can expose incomplete messages.
Should I disable Smart Sending for the test?
First inspect the recorded reason and decide whether the protection is behaving as intended. Use a properly designed authorized fixture rather than weakening live-audience safeguards merely to obtain a send.
Does a successful preview prove the flow works?
No. It helps validate content, but entry, routing, eligibility, and recorded delivery still need separate evidence. Follow one profile through the actual journey.
Fix the first broken condition
For a Klaviyo flow not sending, identify one expected message and one authorized profile. Locate the first point where evidence diverges from the intended journey. Repair that condition, preserve legitimate exclusions, and verify the result before expanding the change to other flows.
Join the conversation
Share a useful perspective, question, or first-hand experience with other Daily Jade readers.