Check each flow against its own last three months before you look at the benchmark

Klaviyo’s benchmarks page will say that a brand’s abandoned cart flow is “fair” for its peer group and below the median on conversion rate. For a monthly review, start with the flow’s own history instead. The benchmark cannot say whether a flow got worse, when, or on which message.

A peer group is about 100 accounts that Klaviyo judges similar to yours in industry, size and scope. The matching uses figures such as average item value, total revenue and year-over-year growth, and campaign characteristics such as the share of days with a campaign sent, taken from the organization’s contact information in Klaviyo. The benchmarks report rates each metric poor, fair, good or excellent over the selected period, against the peer group and Klaviyo’s best practices. When the peer group updates, the historical comparison is recalculated against the new group. An account can be an outlier on one business characteristic. And Apple Mail Privacy Protection inflates open rates, so the open-rate rating is the one to treat with most caution.

The monthly check

The report. In Klaviyo, build a flows performance report grouped by flow and by month, with delivered, click rate, conversion rate (placed order rate, or the account’s chosen conversion metric) and revenue. Revenue per recipient is revenue divided by delivered. Set the range to Last 365 Days and schedule the report to recur monthly, on the third, with an email to you when it’s ready. In flow analytics, Klaviyo attributes email opens and clicks to the date the message was sent, and conversions from late-month sends can still be arriving on the third, so treat the latest month as provisional: next month’s report reads it again as final, and that is the figure the note column records.

The table. One row per flow that delivered at least 200 messages in any of the three months before this one. A flow that delivered nothing this month stays on the table with a zero, which is the row you most want to see. Each cell is this month, the average of the prior three, and the percentage change. In seasonal months, add the same month last year beside the three-month average.

What triggers a closer look. Delivered down by more than a quarter against the three-month average, or revenue per recipient down by more than a third. Those are starting thresholds. Check them against the flow’s own history in the 365-day report: if a flow’s normal month-to-month swing is bigger than the threshold, widen the threshold for that flow until an ordinary month no longer trips it, and write the number in the note column. Delivered comes first because a flow that stopped delivering, from a broken trigger, a changed segment condition, or a message left in draft after an edit, shows up here before it shows up in revenue, and it has nothing to do with the copy.

When a flow trips. Open the flow and read message by message for the month. If the first message’s delivered count dropped, check the trigger and the flow filters before anything else. If one message’s click rate dropped while the others held, open that message and look at what changed in it and when. If every message dropped at once, look upstream: a site change, a stock-out, a season.

The note column. One line per flow per month, written whether or not anything tripped: “flat”, “delivered down 30%, trigger changed on the 12th”, “conversion up, new offer in message 2”.

Where benchmarks come in. Once a quarter, on conversion rate and revenue per recipient, never on open rate. A flow rated poor against its peers and flat against its own history for three months is a candidate for a rebuild; the next step is to read its messages against the best-performing flow of the same type you run for another brand before changing anything. A flow rated poor against peers but improving against its own history is working; keep going and check the rating again next quarter.