How to audit your Klaviyo flows: what's wrong, and what it's costing you
By David Refaeli-Berman
Auditing your Klaviyo flows takes six steps: find where flow revenue is recorded, count which flows are actually sending, compare each one against stores your size, put a dollar figure on every gap, take apart the flow with the biggest gap, then fix it.
Most flow audits end in opinions. This one ends in numbers you can act on.
Disclosure: we make EmailStrategist, which does this automatically. This is our read, not an independent review. Everything below is the manual version, and you can run every step yourself, today, without buying anything.
Five percent of your sends, forty percent of your money
Klaviyo's benchmark data, drawn from more than 183,000 customers, puts automated flows at 5.3 percent of email sends and about 41 percent of email revenue. Per recipient, that's roughly 18 times what campaigns earn.
Nearly half of flow revenue comes from first-time buyers, so flows close first purchases as well as repeat ones.
That gap between 5.3 percent of sends and 41 percent of revenue is why an audit starts with flows: a small flow fix can outperform a much bigger campaign push.
Treat 41 percent as context rather than a target, because it's an average across Klaviyo's whole dataset. Step 3 compares you against stores your size, which is the better test.
Our campaign versus flow guide covers how to read that split for your own store.
Step 1: Find where flow revenue actually hides
Figure out which flows earn the most, because that ranking decides which one you fix. Open the Flows list and sort by the Conversions column, but two settings decide whether that column shows revenue at all.
Set the conversion metric to Placed Order, or anything else with money attached. On a signup event, the column counts signups, not dollars.
Set the metric period to Last 30 days, or a custom range up to 180 days. It defaults to 7, and one week tells you almost nothing about a flow.
The list stops there. For the rest, open Analytics and record four numbers per flow: revenue, revenue per recipient, conversion rate, and how each moved against the month before.
Watch for one pattern: revenue rising while revenue per recipient falls. The flow is reaching more people and converting fewer of them, so the chart climbs while each new entrant is worth less than last month.
Step 2: Count what's live, then what's actually sending
Most checklists tell you to confirm your core flows are live, and that instruction loses money.
Five flows carry the load: welcome, abandoned cart, browse abandonment, post-purchase, and win-back.
Most stores run one abandoned cart flow that covers both cart and checkout. Splitting checkout into its own sequence catches people closer to buying, not a sixth flow to add later.
Three more earn their place once those five are working: replenishment, back-in-stock, and sunset. Our essential flows guide covers all eight, but an audit starts with the five above.
Check each one twice. Is it built and switched on, and has it sent to anyone in the last 30 days?
A flow can be on and dark. A trigger can stop firing after an integration change, a filter can be tightened until nobody qualifies, or a renamed metric can quietly disconnect. In the Flows list it still reads live while in Analytics its recipients sit near zero.
Nothing alerts you, because nothing broke loudly. A dark flow is a missing flow, so count it that way.
The three abandonment flows aren't worth the same either. Each catches a different moment, so rank them by how close the shopper got to buying.
| Flow | Intent | Who it catches |
|---|---|---|
| Checkout abandonment | Highest | Started checkout, didn't buy |
| Cart abandonment | Mid to high | Added to cart, never reached checkout |
| Browse abandonment | Lower | Viewed a product, never added to cart |
Build them in that order, which is how Klaviyo ranks them in its own checklist. If your cart flow doesn't treat checkout abandoners separately, that split is usually the cheapest revenue on this list.
Many guides count "win-back or sunset" as one flow, but they do opposite jobs: win-back tries to earn another order, while sunset stops mailing people who went quiet so your email keeps reaching inboxes. Count them as one and you'll hide whichever one you skipped.
Step 3: Compare each flow to stores your size
Most articles quote one average revenue per recipient for every flow at every brand, and that number can't tell you much about your flow. A $30 order value and a $300 order value don't produce comparable results, and neither do a $500,000 store and a $15 million one.
Klaviyo publishes better numbers than the averages people quote from it. Its help center carries flow benchmarks split by annual revenue, then by average order value, with the 25th and 75th percentile for each.
Here's what it looks like for stores doing $1 million to $5 million a year:
| Your AOV | Abandoned cart | Welcome | Post-purchase |
|---|---|---|---|
| $37 to $44 | $0.71 to $1.68 | $0.29 to $1.73 | $0.06 to $0.23 |
| $44 to $83 | $1.12 to $3.03 | $0.35 to $2.53 | $0.11 to $0.44 |
| $83 to $112 | $2.39 to $5.86 | $0.51 to $3.93 | $0.24 to $0.72 |
| $112 to $163 | $3.86 to $8.63 | $1.15 to $8.39 | $0.29 to $1.30 |
Revenue per recipient, 25th to 75th percentile, from Klaviyo's published flow benchmark reference, accessed August 2026. The full tables include the under-$1M and $5M to $20M bands.
Find your row. Below the 25th percentile means you're behind stores like yours, which is worth a look; above the 75th, leave it for later unless something else says otherwise.
Klaviyo also shows a comparison under Benchmarks, against a peer group rather than the whole database.
Step 4: Put a dollar figure on every gap
An audit that ends in adjectives gives you nothing to choose between, so turn each finding into money first.
For a flow running below its band, its row in the benchmark table above, take last month's recipients, multiply by the 25th percentile for your band, and subtract what the flow actually earned. That difference is the monthly gap.
When a flow is missing or dark, estimate who would qualify each month, then multiply by the 25th percentile for your band. For the cart flow, that's people who add to cart and don't buy; for welcome, it's new subscribers.
Use the 25th percentile rather than the 75th, because you're sizing a gap and not pitching an upside. And remember that attributed revenue isn't the same as revenue you'd have lost, which our attributed revenue guide covers in detail.
Step 5: Take apart the flow with the biggest gap
Stop auditing everything. Open the flow with the biggest dollar gap and check seven things.
The trigger. Flow triggers decide who enters, so start here. Is it firing on the right action, and is that action coming from your best data source? For Shopify stores, an event coming straight from Shopify beats one piped in through an integration, which beats one from an outside app. Confirm it's still firing, and confirm no second flow competes for it.
The filters. Flow filters decide who stays in. Are recent buyers excluded, so customers stop getting cart reminders, and are the conditions so tight that almost nobody qualifies? That's how live flows go dark.
The timing. Does the first send match intent? Klaviyo's guidance for abandoned cart is 30 to 45 minutes, so test it rather than assuming it.
The channels. If you have text consent and the flow is email only, that's another channel worth testing.
Smart Sending. This one hides in plain sight, because it skips anyone who already got an email inside a set window, and the default is 16 hours. Campaigns count toward that window, so in a heavy campaign week your best flow emails can get skipped. Skipped messages are not resent either: Klaviyo does not queue them for later, so that revenue is simply gone. Check the setting on every core flow.
The splits. One good split usually beats one more email, whether that's first-time buyer versus repeat, or high cart value versus low.
The content. One message per email, one obvious button, and an offer you've tested rather than inherited.
Step 6: Rank the fixes and work the list
You now have gaps with dollar figures. Rank them three ways: how much revenue it moves, how hard the fix is, and whether it matters past this month. Start with the biggest gap that's cheap to fix.
Not everything on the list deserves a fix, because some flows should be killed. If two flows chase the same moment they compete for the same person, and Smart Sending decides which one wins, so merge them. If a flow has run for a year and earns almost nothing per recipient, turning it off beats rebuilding it. Fewer flows, better built, usually earn more.
Write down the flow, the change, the reason and the date. A fix without a date is a fix that doesn't happen.
Then book the next audit. Klaviyo recommends refreshing core flows two to three times a year, and if you haven't touched a flow in twelve months you should assume it needs work. Offers change, the same emails stop landing, and the people entering that flow today aren't the ones who entered two years ago.
Klaviyo's own audit button, and what it can't see
There's an Audit flows button on the Flows page. It runs Klaviyo's AI across your flows, segments, and forms, finds an opportunity, and can build the campaign to capture it. Start there, because it's faster than working down any checklist by hand.
Then check the one thing it doesn't look at. Klaviyo's own published audit prompt tells the AI to include only flows marked live and skip everything else. But a dark flow, switched on and sending to nobody, still counts as live. So it never reaches the audit as a problem, which is exactly the gap in Step 2.
Composer runs on credits, so auditing a big account costs money, and it recommends changes rather than making them.
Three jobs this audit skips, and where they live
Deliverability decides whether any of this reaches an inbox, so watch spam complaints, bounces, and unsubscribes.
Campaigns are the other 95 percent of your sends. They're judged on how often you send and who you send to, not on triggers.
List growth limits all of it, because it decides how many people can enter a flow at all.
Fix flows first: revenue per recipient is usually highest there.
The Klaviyo flow audit checklist
- Revenue.
- Set the conversion metric to Placed Order.
- Set the metric period to Last 30 days.
- Sort by Conversions.
- Record revenue, revenue per recipient, conversion rate, and trend.
- Flag any flow with rising revenue and falling revenue per recipient.
- Coverage.
- Confirm welcome, cart, browse, post-purchase, and win-back are built and on.
- Confirm each sent to someone in the last 30 days.
- Check whether a sunset flow exists separately.
- Comparison.
- Find your revenue and average order value band.
- Compare each flow's revenue per recipient to the 25th and 75th percentile.
- Check the peer view under Analytics, Benchmarks, Flows.
- Sizing.
- Turn every gap into a monthly dollar figure at the 25th percentile.
- Diagnosis.
- On the biggest gap, check trigger, filters, timing, channels, Smart Sending, splits, content.
- Plan.
- Rank by revenue moved, how hard the fix is, and whether it matters past this month.
- Give each fix a date.
- Set a reminder for two to three times a year.
The same audit, without spending the afternoon
All of this takes a few hours on a small account and most of a day on a big one. The steps aren't hard, but finding the numbers and lining them up is tedious, and the answer changes every month.
That's the job EmailStrategist does. It reads your Klaviyo the same way this checklist does, flow by flow against your revenue band, then builds the highest-value email you aren't sending and stages it as a draft in your account. You review it. You ship it.
Senior strategist judgment, without the monthly retainer.
One recent audit of an 8-figure apparel brand found a $200,000-a-month gap between the flows they had built and the flows actually sending, sized the same way as Step 4: qualifying recipients at the 25th percentile for their band.
Get your free audit and you'll have your own numbers before you spend an afternoon collecting them.
Is it safe to connect Klaviyo?
- Read-only. It can read your campaign and flow performance. It cannot send, edit, or delete anything in your account.
- No customer data stored. It works from your metrics, not your subscriber list. Names, emails, and phone numbers are never stored.
- Nothing ships without you. Built emails land as drafts for your review; nothing is pushed to Klaviyo on its own. Reading and staging are separate permissions: the audit runs on read-only access, and drafts are staged only if you later grant write access, which still cannot send.
- Built by an operator. David Refaeli-Berman ran enterprise email at Chewy, Amazon, Overstock, and Bed Bath & Beyond.
Questions
How do you audit Klaviyo flows?
Set the conversion metric to Placed Order and the metric period past its 7-day default, then pull revenue, revenue per recipient, conversion rate, and trend for every flow. Confirm your five core flows are switched on and actually sending. Compare each flow to Klaviyo's published benchmark for your revenue and order-value band. Turn every gap into a dollar figure at the 25th percentile. Then fix the biggest cheap gap first.
Where do I find flow revenue in Klaviyo?
On the Flows list, but only once two settings are right. Set the Conversion metric to Placed Order, or another metric with money attached, and the Conversions column shows revenue attributed to each flow. Set the metric period longer than its 7-day default. Then sort by that column. For revenue per recipient and trends, switch the toolbar from List to Analytics, or open the flow reports.
What is a good revenue per recipient for a Klaviyo flow?
It depends on your annual revenue and your average order value, which is why global averages mislead. Klaviyo publishes 25th and 75th percentile figures by band. For a $1 million to $5 million brand with an AOV of $44 to $83: abandoned cart runs $1.12 to $3.03, welcome $0.35 to $2.53, post-purchase $0.11 to $0.44. Below the 25th percentile for your band is a signal worth investigating.
How many flows should a Klaviyo account have?
Start with five core revenue flows: welcome, abandoned cart, browse abandonment, post-purchase, and win-back. Treat sunset as a separate deliverability job. After those, add flows that fit your buying cycle, like replenishment or back-in-stock. Splitting checkout abandonment out of your cart flow is a refinement of a core flow rather than a new one.
Why does a flow show as live but earn nothing?
Because live and sending are different states. A trigger stops firing after an integration change. A filter gets tightened until nobody qualifies. A renamed metric quietly disconnects. The Flows list still reads live while recipients sit near zero. Check recipients in the last 30 days, not the status toggle.
Can Klaviyo audit my flows automatically?
Yes. The Flows page has an Audit flows button that runs Klaviyo's AI across your flows, segments, and forms and returns recommendations, and Composer can build a campaign from what it finds. It recommends changes. It doesn't make them. And it runs on credits. Worth knowing: Klaviyo's own published audit prompt tells the AI to look at live flows only, so a flow that's switched on but no longer sending still counts as live.
Which Klaviyo flows should I turn off?
Turn off a flow when it competes with a better one, or when it has run long enough to prove it earns almost nothing per recipient. Two flows chasing the same moment split the same audience, and Smart Sending decides which one lands. Merge those. An underperforming flow that has been live a year and sits far below the 25th percentile for your band is usually worth killing rather than rebuilding.
What is the difference between a win-back flow and a sunset flow?
A win-back flow tries to earn another order from a lapsed customer. A sunset flow stops mailing people who went quiet, to protect inbox placement. Guides often count them as one slot, which hides whichever one is missing. Our sunset flow guide covers how to stop mailing people who went quiet.
How often should I audit my Klaviyo flows?
Run the full audit two to three times a year, which matches Klaviyo's own advice on refreshing core flows. Treat anything untouched for twelve months as needing work. In between, check revenue and coverage monthly to catch flows that stopped sending.
Find out what your flows aren't earning.
A free audit reads your Klaviyo, ranks the emails you aren't sending, and puts an estimated dollar value on each gap. Your fix list starts with numbers instead of opinions.
Get my free audit