A lot of brands still pick campaign audiences ad hoc. Targeting swings between two bad extremes: the same engaged group (or the full list) on every send, or a constantly changing mix of segments even when campaign type, content, and theme stay the same.
Neither gives you a clean read after the fact. This walkthrough shows the alternative we use with clients: a campaign targeting ruleset that feeds a Send Builder — pick a few parameters, get the current target and suppression logic, then update the book instead of reinventing segmentation every send.
One pattern is inertia: every campaign hits the same general engaged group, or worse, the full list. That is simple, and it is rarely ideal.
The other pattern looks more sophisticated but is just as hard to learn from — lots of segments, changed often, even when the campaign theme is consistent. After the fact you cannot tell whether performance moved because of creative, cadence, or because the audience was a different group than last time.
Analysis needs consistency. You cannot iterate a segmentation strategy if the targeting rules themselves keep moving.
The system starts as a living document, not a one-off Klaviyo screenshot. Live vs. draft versions let you iterate without losing what is currently in production. A change log captures why a rule moved, so later analysis has a paper trail.
The point is to get more iterative on how you update the rules — and more consistent on how you actually target — instead of guessing in the ESP for each campaign.
The front end is a targeting builder. Choose the parameters for that send; it returns the current target segments and suppression segments to copy into Klaviyo (or whatever ESP you use).
Typical inputs: theme (Seasonal, Promotion, and so on), stage (launch, reminder, single send), version (default, or a split like men’s / women’s), and reach (tighter vs. wider). The output is the audience logic for that combination — not a blank-slate debate in Slack.
Behind the builder is an engagement ladder — volumes and tiers of engagement groups — seeded from the past year of sends. For example: Seasonal single-sends went out 129 times (~10.8/month); 35% of those included T2, 21% T1, 21% T3; suppressions split into baseline, regional, and other types.
That table can get large. Phase one is not a fancy model — it captures current practice, distills a default when you are not going wider or tighter on purpose, and powers the builder so coordinators can run by the book.
Analytics layer on later: which segments actually work, then a new version of the logic. Until then, the job is to stop going ad hoc.
Once the builder is the source of truth, a marketing coordinator can take it and send. Strategy bandwidth goes to analysis and to shipping updated versions of the rules when the data supports it — not to reinventing targeting every campaign.
Learn more: A theme-based approach to email campaign calendars, Retention marketing strategy, and Book a call.
Prefer YouTube? Watch on YouTube