The Flow Audit: Where Your Automations Leak Revenue

Why do email flows leak revenue?

Most flow problems are not one dramatic failure. They are a stack of small logic, consent, timing, content, and QA gaps that compound over time. ZHS uses a 61-point internal audit; the public version below includes the core checks and a downloadable working checklist. Across a built-out program, flows should contribute 30 to 50% of email-attributed revenue, not total store revenue.

The canvas can look complete while customer paths, product logic, consent, reporting, or message pressure still fail. ZHS uses the 61 checks to make those assumptions explicit before a flow is marked approved.

Your flows can be "on" and still leak.

A flow can be live yet still have an unintentional branch, channel role, timing rule, or exit condition.

Here is a sample of what those 61 points check.

---

What does a fully built-out flow look like?

A built-out flow is not a preset message count or an automatic SMS track. It is a customer journey with deliberate trigger, eligibility, message job, channel role, exits, pressure settings, and end-to-end QA. "Set up" can mean a trigger fires; "built out" means the intended path has been tested and documented.

Most brands set a flow live and call it done. One welcome email. One cart reminder. A trigger that fires.

That is "set up." It is not built out.

Built out means the intended sequence, branching, channel role, exit logic, and QA path have been documented and tested end to end.

---

Which email flows should you audit first?

Start with the customer journeys that actually apply to the store, often welcome, site or browse abandonment, cart, checkout, post-purchase, and re-engagement. Do not treat that list, a message count, or an SMS path as universal. Every qualifying path needs intentional trigger, eligibility, exit, and reporting logic. If foundational journeys are missing, start with the complete welcome flow guide and abandoned cart recovery guide.

Before optimization, confirm every relevant core journey exists and passes its functional path. Evaluate the combined flow contribution against email-attributed revenue only.

Flow	Built-out standard	Common leak
Welcome	Documented onboarding sequence with a clear job per message	Single discount delivery without onboarding context
Site Abandon	Use only when tracking and a useful customer action support it	Assumed necessary without viable tracking or message job
Browse Abandon	Product context, eligibility, and reporting treated independently from cart	Merged into cart without testing the behavior difference
Cart Abandon	Own trigger, purchaser exit, and tested recovery treatment	One reminder with no exit or measurement plan
Checkout Abandon	Own trigger and cohort reporting, separate from cart	Assumed to be the same path as cart
Post-Purchase	Lifecycle guidance based on product use, education, or repeat cycle	Thank-you message with no defined next customer job
Winback / Sunset	Re-engagement and suppression policy matched to engagement data	Unengaged profiles never handled

---

What are the most common flow leaks?

This guide checks five recurring leak classes: generic customer treatment, channel duplication or omission, weak timing, missing exits, and dead sends. Each can damage a customer path even when the canvas looks live. The correction is not “add SMS everywhere”; it is to give every channel a consented, measurable job.

Once the flows exist, this is where revenue quietly drains inside them.

1. No segmentation branches. The flow sends the same content to a first-time visitor and a repeat buyer. Splits should follow your email segmentation strategy, not fire one path for everyone.
2. Channel mismatch. A flow either omits a consented, useful channel or duplicates the same message across channels. The SMS flow guide explains how to give email and SMS distinct jobs.
3. Weak timing. Back-to-back sends with no delay, or account-level pressure settings that conflict with the intended journey.
4. No exit criteria. Buyers stay in the abandonment flow and keep getting "you forgot something" after they already paid. Filters should pull them out.
5. Dead sends. Discount trapped in an image instead of live HTML text. Broken auto-apply links. Out-of-stock products named in copy. Any one turns a working send into a dead one.

---

What share of email revenue should flows drive?

Flows should contribute 30 to 50% of email-attributed revenue across a built-out program. Do not assign universal store-revenue percentages to individual flows: welcome, abandonment, and post-purchase performance changes with traffic mix, catalog, repeat cycle, eligibility, and attribution settings.

Use flow-specific rates to diagnose the build: welcome first-purchase rate and total attributed revenue, cart and checkout recovery, post-purchase repeat rate, and winback conversion. Then use the combined flow share to judge whether automation carries enough of the email program.

---

How do you prioritize flow fixes?

Score each gap by impact and effort, then work high-impact, low-effort first. In practice that means four steps: cover missing customer journeys, fix broken sends, add appropriate structure, then test and review on a documented cadence. You cannot resolve every check across every live journey at once, so sequence the work.

Score each gap by impact and effort. Then work high-impact, low-effort first.

1. 01
   Missing journeysCover the customer path

   Build the relevant journey before polishing a lower-impact message.

2. 02
   Broken sendsFix dead links and product logic

   Bad auto-apply links, image-only codes, and unavailable products break an otherwise valid customer path.

3. 03
   StructureAdd deliberate branches, channels, and exits

   Use consented channels, segment logic, timing, and purchase exits where they change the customer experience.

4. 04
   OngoingTest and review on a recorded cadence

   Compare delivery, clicks, conversion, total attributed revenue, and skip reasons under stable reporting definitions.

Structured from the canonical article steps for responsive reading and presenter mode.

---

How do you run the audit inside Klaviyo?

Start in Flows, filter to Live and Manual statuses, and inventory every automation before opening individual messages. For each flow, record its trigger, profile and trigger filters, re-entry behavior, live/manual message count, channel, last edit, and performance over a documented comparison window. Then use a test profile to force the trigger and inspect the event timeline rather than assuming the canvas logic works.

1. 01
   InventoryMap every flow and status

   Export or record Live, Manual, and Draft flows. Flag duplicate triggers and missing core journeys.

2. 02
   LogicInspect triggers, filters, splits, and exits

   Confirm eligibility, purchaser exits, re-entry rules, consent checks, and mutually exclusive branches.

3. 03
   ExperienceTest every message as a customer

   Force each trigger with test profiles, receive the messages, click every link, apply every code, and verify mobile rendering.

4. 04
   PerformanceCompare like with like

   Use the same date window and attribution settings. Review delivery, click, conversion, total attributed revenue, and skip reasons by message.

5. 05
   PrioritizeAssign owner, impact, and due date

   Fix broken logic first, missing coverage second, then weak creative and incremental tests.

ZHS default: run this pass monthly for business-critical flows and after any catalog, offer, integration, or consent change. Adapt the cadence to the actual risk and change rate.

• ✓Trigger fires from a real storefront action
• ✓Trigger filters include only the intended event
• ✓Profile filters exclude ineligible or non-consented contacts
• ✓Purchasers exit every abandonment path
• ✓Re-entry rules prevent repetitive enrollment
• ✓New and returning customer branches are intentional
• ✓Email and SMS consent are checked separately
• ✓Quiet hours and local time behavior are verified
• ✓Smart Sending behavior matches the journey design
• ✓Delays do not cause messages to collide
• ✓Every message is Live or intentionally Manual
• ✓Subject, preview, From name, and reply-to are correct
• ✓Dynamic product and profile variables have fallbacks
• ✓Discount codes and auto-apply links work
• ✓Every CTA lands on the intended mobile page
• ✓Out-of-stock and excluded products are handled
• ✓UTM and campaign naming follow the reporting standard
• ✓Test profiles cover eligible, ineligible, buyer, and non-buyer paths
• ✓Skip reasons and delivery failures have been reviewed
• ✓Every issue has an owner, priority, and due date

A flow is not approved because its canvas looks complete; it is approved when the customer path and reporting both pass.

Download the complete 61-point flow-audit checklist (CSV) →

What are the most common flow audit mistakes?

The six recurring mistakes are calling a flow done at launch, leaving pressure settings unexamined, adding or skipping channels without a reason, missing purchase exits, trapping a code inside an image, and never re-auditing. Each is testable with a controlled profile and a documented review pass.

1. Calling a flow done at launch. A live flow still needs the full 61-point pass before it is marked approved.
2. Never checking pressure settings. Smart Sending and other account-level controls must match the journey design and be tested with a real profile.
3. Treating SMS as mandatory or optional by default. Add it only when valid consent, pressure rules, and a distinct customer job support it.
4. No purchase exit filter. Customers who bought keep getting abandonment reminders. That erodes trust fast.
5. Discount trapped in an image. Put the code in live HTML text and confirm the auto-apply link works.
6. Never re-auditing. Flows drift after product, offer, integration, consent, and catalog changes. ZHS defaults to a monthly business-critical-flow review and a review after material changes; adjust that cadence to the operating context.

---

Get Expert Help

Our team runs the full 61-point flow audit as part of onboarding, documents priority gaps, and sequences fixes by expected customer-path impact.

You get built-out flows, not just flows that are technically switched on.

See our pricing | Apply to work with us