Use Klaviyo's Predictive Analytics to Build High-Value Segments

What This Does

Klaviyo's ML models score every subscriber on predicted lifetime value, churn risk, and expected next order date — so you can build segments like "customers likely to churn this month" without any data science skills.

Before You Start

• You have a Klaviyo account with at least 3-6 months of purchase history synced
• Your e-commerce platform (Shopify, WooCommerce, etc.) is integrated with Klaviyo
• You're logged in to Klaviyo

Steps

1. Navigate to Lists & Segments

In the left sidebar, click Audience then Lists & Segments. Click the Create List / Segment button in the upper right. Choose Segment (not List — segments are dynamic and auto-update as customers move in and out).

2. Add a predictive analytics condition

In the segment builder, click Add a condition. In the category dropdown, scroll to or search for Predictive analytics. You'll see options including:

• Predicted CLV (customer lifetime value) — low, medium, high, or very high
• Churn risk — low, medium, or high
• Expected next order date — within X days

3. Build a churn-risk segment

Select Churn risk > is > High. Add a second condition: Person > has placed at least 1 order. This gives you high-value customers who Klaviyo predicts are about to stop buying — your retention target list.

What you should see: A preview count of subscribers in this segment updates in real time as you add conditions.

4. Save and name the segment

Give it a clear name: "High Churn Risk — Active Buyers." Click Create Segment. Klaviyo will build and maintain this segment automatically — it updates as churn risk scores change.

5. Use the segment in a campaign or flow

Go to Campaigns and create a new campaign. In the audience selection step, choose your new "High Churn Risk" segment. Write a targeted re-engagement campaign just for this group — with a compelling offer, since these are customers Klaviyo predicts may not come back.

What you should see: Your campaign audience shows a smaller, targeted list instead of your full subscriber base.

Troubleshooting: If predictive analytics options aren't showing, you may need more purchase history. Klaviyo requires sufficient transaction data to generate predictions — this is a paid plan feature.

Real Example

Scenario: You manage email for a subscription coffee brand. You want to proactively retain subscribers before they cancel.

What you type/do: Build segment: Churn Risk = High AND placed at least 2 orders. Result: 847 subscribers.

What you get: A list of 847 previously active customers who Klaviyo predicts are about to lapse. You send them a personalized "We miss you" campaign with a 15% discount — recovering those who would have churned silently.

Time saved: Replaces what would otherwise be a manual data pull and analysis exercise, delivered instantly.

Tips

• Combine predictive segments with behavioral conditions for precision: "High churn risk AND hasn't clicked an email in 60 days" is much more targeted than either condition alone.
• Predicted CLV segments are especially powerful for VIP programs — build a "High Predicted CLV" segment and treat them differently: early access, better offers, personal check-ins.
• Check segment sizes regularly. If your High Churn Risk segment is growing month over month, it signals a larger retention problem worth investigating.

Tool interfaces change — if a button has moved, look for similar AI/magic/smart options in the same menu area.