Klaviyo Flows the Operator Watches Weekly

Q: What if a flow metric is slightly off but not flagged?

If a metric is within the threshold but trending in the wrong direction (e.g., open rate at 18% but down from 22% last month), add it to the yellow list and monitor for 2 more weeks. Trends matter more than single-week anomalies.

Q: Should we pause a flow if one metric flags?

No. Pause only if delivery rate drops below 85%, unsubscribe rate exceeds 2%, or if the flow is sending to the wrong segment. Otherwise, investigate and fix while the flow runs.

Q: How do we know if a segment definition is wrong vs. the segment is just smaller?

Compare segment size to the same week last year and the same week last month. If size is stable month-over-month, the segment is fine. If size dropped 50% in one week, the definition likely broke.

Q: What's the difference between a bounce and an unsubscribe?

A bounce is a failed delivery (hard bounce = invalid address, soft bounce = mailbox full). An unsubscribe is a customer clicking the unsubscribe link. Both hurt performance but require different fixes - bounces need list cleaning, unsubscribes need content review.

Why Weekly, Not Daily

Daily monitoring creates alert fatigue. Weekly review allows operators to distinguish signal from noise - a single failed send is noise; three consecutive days of low open rates is signal.

Most Klaviyo flows run on predictable cadences (welcome series, post-purchase, cart abandonment, re-engagement). Weekly review aligns with the natural rhythm of these campaigns and gives enough data to spot trends.

The cost of missing a flow break for 7 days is typically 2-5% of weekly revenue from that flow. The cost of checking daily is 3-4 hours of operator time per week. Weekly review is the operational sweet spot.

The Five Flows to Watch

Not all flows are equal. These five carry the highest revenue density and break risk:

Welcome series (segment: all new subscribers) - breaks here affect 100% of new customer LTV
Post-purchase / order confirmation (segment: all orders) - breaks here create support tickets and refund requests
Abandoned cart (segment: cart value > $0, no purchase in 24h) - breaks here leak 8-12% of weekly revenue
Browse abandonment (segment: product view, no purchase in 48h) - breaks here affect mid-funnel re-engagement
Win-back / re-engagement (segment: no purchase in 90+ days) - breaks here silently degrade list quality

Weekly Checklist: Metrics to Review

Each flow requires a 5-minute review. Check these metrics in order:

Delivery rate: Is it 95%+? If below 95%, check for hard bounces, list decay, or ISP blocks. Threshold: flag if week-over-week drop > 3%.
Open rate: Is it within 10% of the 4-week average? Threshold: flag if open rate drops > 15% from baseline.
Click rate: Is it within 10% of the 4-week average? Threshold: flag if click rate drops > 20% from baseline.
Unsubscribe rate: Is it below 0.5%? Threshold: flag if unsubscribe rate > 1% in a single send.
Segment size: Is it growing or stable? Threshold: flag if segment shrinks > 20% week-over-week without explanation.
Bounce rate: Is it below 2%? Threshold: flag if bounce rate > 3% or if hard bounces spike.

Common Break Patterns and Fixes

Breaks rarely appear as total flow failure. They appear as subtle metric shifts. Recognize these patterns:

Segment stopped growing (welcome series): Check if form integrations are still firing. Verify Shopify app permissions. Test form submission end-to-end.
Delivery rate dropped (any flow): Check if domain reputation degraded. Review recent email content for spam trigger words. Verify DKIM/SPF records.
Open rate collapsed (any flow): Check if send time changed. Verify subject line wasn't accidentally templated with broken variable. Check if list was recently imported with bad data.
Click rate dropped but opens stable (any flow): Check if links were recently changed. Verify UTM parameters aren't malformed. Test link clicks in email client.
Unsubscribe spike (any flow): Check if email frequency increased. Review recent subject lines for misleading claims. Verify CTA copy matches email content.
Bounce rate spiked (any flow): Check if list was recently merged with external data. Verify no bulk import of unverified emails. Review for typos in email addresses.

Remediation Decision Tree

When a metric flags, follow this sequence:

Step 1: Confirm the break is real. Check if the metric shifted in the last 7 days. If it's a single anomalous send, wait for next send to confirm trend.
Step 2: Isolate the cause. Check flow settings, segment definition, email content, and integrations in that order.
Step 3: Implement fix. Do not pause the flow unless delivery rate is below 85% or unsubscribe rate exceeds 2%.
Step 4: Monitor next send. Verify the fix worked before closing the ticket.
Step 5: Document. Log the break, cause, and fix in a shared operator log for pattern detection.

Segment Health Verification

Segment breaks often precede flow breaks. Review segment definitions weekly for these failure modes:

Check if any segment uses a custom property that stopped updating. Example: 'last_purchase_date' stops updating if Shopify integration breaks.

Verify segment logic hasn't drifted. Example: 'cart value > $0' may exclude orders under $1 if a recent product launch added sub-$1 items.

Confirm segment size is within expected range. If a segment that normally has 5,000 members drops to 500, the segment definition likely broke.

Test segment membership manually. Pick 3 random customers who should be in the segment and verify they actually are.

Documentation and Escalation

Maintain a weekly flow report with three sections: green (all metrics normal), yellow (one metric flagged, fix in progress), red (multiple metrics flagged or delivery < 85%).

Escalate to the email specialist if: delivery rate drops below 85%, open rate drops below 10%, or a segment stops growing for 2+ weeks.

Escalate to the developer if: a custom property stops updating, an integration stops firing, or a segment definition uses logic that can't be verified in the UI.

Escalate to the brand if: unsubscribe rate exceeds 1.5% or a flow needs to be paused due to content issues.

Questions

FAQ

What if a flow metric is slightly off but not flagged?

If a metric is within the threshold but trending in the wrong direction (e.g., open rate at 18% but down from 22% last month), add it to the yellow list and monitor for 2 more weeks. Trends matter more than single-week anomalies.

Should we pause a flow if one metric flags?

No. Pause only if delivery rate drops below 85%, unsubscribe rate exceeds 2%, or if the flow is sending to the wrong segment. Otherwise, investigate and fix while the flow runs.

How do we know if a segment definition is wrong vs. the segment is just smaller?

Compare segment size to the same week last year and the same week last month. If size is stable month-over-month, the segment is fine. If size dropped 50% in one week, the definition likely broke.

What's the difference between a bounce and an unsubscribe?

A bounce is a failed delivery (hard bounce = invalid address, soft bounce = mailbox full). An unsubscribe is a customer clicking the unsubscribe link. Both hurt performance but require different fixes - bounces need list cleaning, unsubscribes need content review.

Want this on your account?

Thirty minutes. Bring the number that keeps you up.

Book a demo