How to read a Klaviyo flow report
Read a Klaviyo flow report in the right order: recipients and skips, opens and clicks, orders and the attribution window, then bounces, spam and unsubscribes.
What this is for
A flow report answers two questions: is each email reaching the people it should, and is it earning its place in the sequence. Most people open it, see an open rate and stop. The more useful numbers are further down: who was skipped and why, how many orders each email is credited with, and how many people it annoyed into unsubscribing.
Klaviyo's flow analytics show every message in a flow side by side, so you can compare email 1 with email 2 on the same measures. source, checked 28 September 2026
Time needed and what you need
Time needed: about 30 minutes per flow. Ten minutes to set the date range and conversion metric, the rest reading each email's report and writing down what to change.
- An Owner, Admin, Manager or Analyst role in Klaviyo, which is required to see flow analytics
- A flow that has been live for at least a few weeks
- Your average order value, to sense-check the revenue figures
- A note of any changes made to the flow recently, such as new filters or delays
The steps
Read the report top to bottom in this order. Each step explains the number before it, so skipping ahead leads to wrong conclusions.
1. Turn on analytics inside the flow builder
Open the flow and switch on the analytics view in the builder's header. Every email card expands to show its own numbers, and you can set the time range for all of them at once. Pick a range long enough to include at least a few hundred recipients per email; a week of data on a small flow is noise. source, checked 28 September 2026
2. Check which conversion metric the report is using
The conversion column counts whatever metric is selected, which is Placed Order unless someone changed it. For a store flow, orders are what you want. If you switch it on one flow email, the change carries across every flow email's report, so set it back afterwards if you were only looking. source, checked 28 September 2026
3. Read recipients and skipped before anything else
Recipients is how many people the email was actually sent to. Skipped is how many people reached that email but did not get it. A large skipped number is not automatically a problem: Klaviyo re-checks the flow filters before each send, so in a flow that stops when someone buys, every buyer shows up as skipped. source, checked 28 September 2026
Open the recipient activity tab and look at the skip reasons. Smart Sending, suppressed profiles and failed filters are normal. A coupon that no longer exists or an out-of-stock product block is a fault you need to fix.
- Normal: failed flow filters, Smart Sending, suppressed profile
- Fix now: missing coupon code, out-of-stock product, no email address
4. Read open and click rates as unique rates
Flow open and click rates count each person once, however many times they opened. Clicks are the steadier signal: Apple Mail Privacy Protection loads images automatically, which counts as an open, so open rates run high and are best used to compare one email with another rather than as a verdict. source, checked 28 September 2026
5. Read orders and revenue with the attribution window in mind
An order is credited to an email when the buyer opened or clicked that email within the attribution window before buying. By default that window is five days for email. So credited revenue means "bought within five days of engaging", not "bought because of this email". Someone who would have bought anyway still counts. source, checked 28 September 2026
Divide each email's credited revenue by its recipients to get revenue per recipient. It is the fairest single number for comparing emails, because it accounts for how many people each one reached.
6. Read the deliverability columns
Below the engagement graph sit delivery, bounce, spam complaint and unsubscribe rates. Klaviyo gives a specific pair of numbers that means your list needs cleaning. source, checked 28 September 2026
The spam complaint rate is the share of recipients who clicked "report spam", not the share that landed in spam. An email with a rising complaint rate is usually one sent too often, too late or to the wrong people. source, checked 28 September 2026
7. Look at links to see what people wanted
The link activity tab lists each link with its unique clicks. If most clicks go to the logo or the footer rather than the main button, the button is not doing its job. If people click a product lower down, move it up. source, checked 28 September 2026
8. Compare emails and write down one change
Line the emails up by revenue per recipient and unsubscribe rate. An email that earns little and loses many subscribers is the one to rewrite or remove. An email that earns well but sits after a long delay may deserve an earlier slot. Write down one change per flow, make it, and read the report again in a month.
A worked example
For a candle store reviewing 30 days of its two-email browse abandonment flow
Worked example (an invented store, with invented numbers)
Email 1 went to 2,400 recipients, with 600 skipped for failing the flow filters. It shows 72 credited orders at an average of $48, so $3,456 in credited revenue, or $1.44 per recipient ($3,456 / 2,400).
Email 2 went to 1,900 recipients and shows 38 orders, $1,824 in credited revenue, or $0.96 per recipient ($1,824 / 1,900). Its unsubscribe count is 41, against 12 for email 1.
Email 1 also shows 30 bounces among its 2,400 recipients, a bounce rate of 1.25% (30 / 2,400), which is over Klaviyo's 1% line. The store makes two changes: it runs its list-cleaning segment before the next campaign, and it tests email 2 with a longer delay to see if the unsubscribes fall.
Common mistakes
Reading the 30-day snapshot in the sidebar as the whole story. It is a quick look at the last month only. Open the full report for the message and set a longer range before drawing conclusions. source, checked 28 September 2026
Treating credited revenue as extra revenue. Attribution counts anyone who engaged within the window, including customers who were going to order anyway. For a true measure of lift, hold back a random group from the flow and compare.
Comparing this month's numbers with last month's without checking the dates. Opens and clicks are counted on the day the email was sent, so an email sent on the 30th that is opened on the 2nd belongs to the earlier month. source, checked 28 September 2026
Changing three things in a flow at once, then reading the report. If revenue moves, you will not know which change did it. Change one thing, or use a flow A/B split.
Questions people ask
Why does my flow show far more people skipped than received?
Usually because the filters are doing their job. In an abandoned checkout or browse flow, everyone who buys before the next email is skipped. Check the skip reasons in the recipient activity tab; if they are filter failures, the flow is working. The guide to setting up an abandoned cart flow explains that filter.
What does Needs Attention mean on the deliverability tab?
It flags an email domain or country where two or more deliverability measures are poor, and only when that domain or country makes up a meaningful share of the email's sends. Look at which provider it is before changing the email itself. source, checked 28 September 2026
Should I change the attribution window to make email look better?
No. Pick the window once, based on how long your customers take to decide, and leave it. Changing it rewrites past reports too, which makes month-to-month comparison meaningless. If you want to report results to someone else, see reporting marketing results to a client. source, checked 28 September 2026
See this on your own store
Paste your store URL. The first pass takes about a minute and needs no account. Save a card (nothing is charged) and Tilly reads the whole store and works out who buys from you.