Flows open at around 32% and click at around 5.6%. Campaigns open at around 31% and click at 1.7%. In clothing, a flow places an order for about 1.6% of the people it reaches against 0.97% for a campaign, and across the whole Klaviyo base a flow earns roughly $1.94 per recipient where a campaign earns $0.11. Those are the 2026 numbers, they are about right, and holding your account against them will mislead you more often than it helps.
Where the numbers actually come from
Search for them and you land on a page that sells the report rather than showing it. The 2026 edition is built on more than 196,000 brands, and the figures everyone quotes are secondhand from it. A median across 196,000 brands spanning every category, price point and list size is a fair description of email in general. It is a weak reference for one store, in the way that the average height of a country tells you nothing useful about whether your own child is growing.
That is the ordinary complaint about benchmarks and it is the least of the problem. The larger one is that the metric they publish is not stable, even inside a single account.
The denominator moves when you change the flow
Klaviyo computes a flow's conversion rate as orders attributed to the flow divided by recipients, summed across every message in it. A four-email flow therefore carries roughly four times the denominator of a one-email flow. Entry-email conversion rate, the share of people who buy off the first touch, is a different number entirely, and it reads higher by construction.
Follow that through. You add a third email to a cart flow, it earns money, and your flow's printed conversion rate goes down, because the third send put several thousand more recipients underneath the same orders plus a few. Delete two emails from a working flow and the rate climbs while the revenue falls. Anyone judged on that percentage has an incentive to shorten your flows.
So when a table says a cart flow converts at 5%, it has not told you the one thing you need to make the comparison: how many emails were in the flows that produced it. Our answer on why three cart emails beat one makes a flow look worse on the benchmark metric and better in the bank, and that tension is a property of the metric rather than of the flow.
Your account is too small to read the gap
This is the part that gets skipped, and it is arithmetic rather than opinion.
We worked out the detection floor for a real account with a list of roughly 7,000 people. Split for a test, each variant received 3,712 emails, and the smallest difference the data could actually resolve was a 35% relative lift. Anything under that was indistinguishable from chance. The rule it produced for us was to stop testing small things on small lists and test offer, angle, segment and voice instead.
The same arithmetic governs any single month's number. A 1.6% order rate measured on 3,712 recipients carries a 95% band of about four tenths of a point either side, so the honest reading is that the true rate sits somewhere between 1.2% and 2.0%. If your cart flow converted at 1.4% last month and the table says 1.62%, you have not found a problem. You have found the width of your own measurement.
Run it yourself before you act on any gap: the band is roughly two times the square root of your rate times one minus your rate, divided by the number of recipients. On a few thousand sends it is almost always wider than the distance between you and the median.
Who is in the denominator at all
A rate also assumes the flow knows about everyone it should be mailing, and for the highest-value flows it does not.
Started Checkout is a browser event. Ad blockers, privacy settings and a script that failed to load all eat it, so a cart flow can post a healthy recovery rate while reaching a fraction of the people who abandoned a cart that week. The ones it never heard about never entered the denominator, which is why the percentage looks fine while the revenue stays flat. Reconciling the two takes about ten minutes, and the check for a cart flow firing short is the one worth running before you tune a single subject line.
It cuts the other way too. A flow with a tiny, well-qualified trigger posts rates that beat every published figure and earns very little, because the audience is small. Browse abandonment usually lands exactly there.
What to compare against instead
Your own account, three months ago, at the same list size, on the same flow. That comparison holds the denominator, the trigger and the audience steady, which is the only way a two-decimal difference means anything.
Two numbers carry more weight than any rate. Revenue per recipient, because it survives list growth and does not move when you add an email. And entrants per week, because a flow with eleven people entering it is not underperforming against a benchmark, it is a flow without an audience, which is the same confusion behind where email revenue actually comes from.
Use the published figures for the one thing they are good for: telling you that flows earn many times what campaigns earn per recipient, so if you have no flows, build flows. For anything finer, the numbers are not wrong, they are just not about you.
What is a good conversion rate for a Klaviyo flow?
The published answer for a cart flow is a median near 5% with the better accounts quoted between 8% and 12%, and the width of that spread is the useful part. Welcome flows are quoted higher, because the trigger catches people at peak interest. Winback flows sit far lower and should.
Before you read anything into your own figure, check how many emails are in the flow and how wide your measurement band is. A number above the quoted range sometimes means the trigger is too narrow rather than the email being good.
Why did my flow conversion rate fall after I improved the flow?
Almost certainly because you added an email. Every recipient of the new send joins the denominator, while the orders it brings are a small addition on top of what the earlier emails already captured.
Check revenue for the flow over the same window. If revenue went up and the rate went down, the flow got better and the metric got worse, and you should keep the email.
Are flow benchmarks worth looking at at all?
Once a year, as a sanity check on direction. They answer questions like whether flows beat campaigns and whether your open rate is in the right order of magnitude.
They cannot answer whether this flow is underperforming this month, because they do not know your list size, your trigger health or how many emails you have in the sequence. That question is answerable, and the data for it is already in your account.
