Skip to main content
Composer

Segmentation Rebuild: Predict Purchase Behavior, Not Opens

Rebuild Klaviyo segmentation around purchase behavior, not engagement proxies.

Mariel Bacci-Kilroy — COO, Sticky Digital
Paste into Composer or any AI
You are a retention strategist specializing in Klaviyo segmentation architecture. My current segmentation is either the Klaviyo default or something built without a clear framework, and I need to replace it with something that actually predicts purchase behavior instead of measuring email opens. Here is my context: Total email list size: {x} Engaged portion: {x} Unengaged: {x} Average order value: ${x} Average purchase frequency for repeat customers: {x} orders/year Average days between first and second purchase: {x} days Repeat purchase rate: {x}% Product model (e.g., one-time purchase, subscription, both): {x} Top 3 purchase categories or SKUs: {x, x, x} Typical customer journey (e.g., "Most first purchases come from a promotion, second purchase typically happens within 60 days or not at all"): {x} My current segments: {x} Build me the following: A behavioral segmentation architecture: Organize it around what customers have actually done, not how recently they opened an email. I want segments that predict purchase intent. Build the tiers (e.g., prospects, active buyers, at-risk, lapsed, and VIP) using my actual purchase frequency data and repeat purchase rate as the inputs for the thresholds. Don't use generic 30/60/90/180-day windows unless my data supports them. My customer journey description matters here. Use it. Exact Klaviyo filter logic for every segment: Be specific enough that a coordinator can build it without asking a single clarifying question. Use real Klaviyo properties, real operators (e.g., "Placed order count is greater than 1” AND “Placed order date is within last {x} days” AND “Predicted customer lifetime value is greater than ${x}"). I don't want interpretation. I want conditions. Suppression and exclusion logic: Tell me which segments should be excluded from which campaign types and why. Most brands suppress unengaged contacts but forget to suppress recent purchasers from acquisition-style campaigns, loyalty members from general discount campaigns, or VIP customers from mass promotional sends. Give me every exclusion I need to manage, with the Klaviyo filter logic for each. * Send frequency recommendation by segment: Give each segment a different cadence tied to purchase probability and churn risk. Don’t give me one frequency for the whole list. That's what I'm trying to fix. If any of my inputs suggest a data quality issue I should address before building this, tell me. Don't build a sophisticated architecture on top of broken data.
Or open in

Customize these fields

  • email-list-sizeEmail list size
  • AOVAverage order value
  • RPRRevenue per recipient
  • current-segmentsCurrent segments