Klaviyo can predict your next churn. Most brands don't have it turned on.

You send the same batch email to your entire list every Tuesday. A chunk of those subscribers haven't purchased in 90 days. Their open rates are dropping. They're about to go cold. Klaviyo already knows exactly who they are. You're blasting them with the same campaign everyone else gets.

Klaviyo predictive churn scoring has been running on the platform for years. It identifies at-risk customers before they stop buying. Over 151,000 brands are on Klaviyo. Most have never looked at the predictive analytics report, let alone built a segment from it.

Klaviyo predictive churn scoring surfaces the customers most likely to stop buying in the next 30-60 days, letting you fire a targeted reactivation flow before you've already lost them. That's a fundamentally different game than sending weekly batch campaigns and hoping the right person opens.

What Klaviyo predictive analytics actually does

Klaviyo's predictive model ingests purchase history, order frequency, average order value, time since last purchase, and email engagement signals. It produces three scores per customer: churn risk, predicted CLV, and next purchase likelihood. These scores update automatically as behavior changes.

The churn risk score is the most actionable. When a customer who normally buys every 45 days goes 80 days without a purchase and stops opening emails, the model flags them as at-risk. You can segment on that prediction right now. Build a flow that fires when someone enters "predicted to churn" status, before they've made any obvious exit signal.

Predicted CLV tells you which customers are worth investing in. If your top 15% by predicted lifetime value get the same email as everyone else, you're leaving money on the table. Earlier access to launches, higher-tier offers, a different reactivation incentive — those decisions start with knowing which segment you're working with.

Smart send time: individual optimization, not batch guessing

Most brands pick a send time once. Monday at 10am. Tuesday at 7pm. Whatever felt right when they set it up. That send time applies to every subscriber on the list.

Klaviyo's smart send time optimization works differently. It routes each email to the window when your specific subscriber is most likely to open, not when the average of your audience opens. One customer might get the email at 6:30am because that's when they check their inbox. Another gets it at 8pm. The send time is per-person, not per-campaign.

The result is a 5-12% open rate lift across the brands that have activated it. The calibration takes 2-4 weeks of data before the model settles. Most brands try it, see no lift in week one, and turn it off. That's the wrong call. The model needs time to learn individual patterns.

The retention math behind predictive churn scoring

A fashion retailer using Klaviyo's predictive features hit a -22% churn rate reduction and recovered $180K in revenue through a single reactivation push. A beauty brand running predictive-powered personalization reached +35% revenue per recipient while sending 18% fewer total emails.

Those numbers come from targeting fewer people more precisely, not blasting the whole list and waiting for someone to convert.

Your list degrades every year. Customers go cold, unsubscribe, or stop buying. If you're not identifying and reactivating at-risk customers proactively, you're losing revenue base while paying to acquire new customers to replace the ones you're bleeding out. The acquisition cost for a new customer is always higher than the cost of keeping one you already have. This is why most DTC brands leave 30-40% of retention revenue on the table — the system doesn't run, not because the tools aren't there.

Why most brands never activate this

Two reasons. Both are fixable.

First, the thresholds. Predictive analytics requires 500+ customers with 180+ days of purchase history. New brands and smaller DTC stores that haven't cleared both requirements don't get access to the model. This isn't a bug — there's not enough data to run reliable predictions below that scale.

Second, the setup friction. Predictive analytics and smart send time aren't on by default. You have to build the segments, connect them to flows, and wait through the calibration period before you see results. Most brands set up a welcome flow, maybe an abandoned cart sequence, and stop there. The advanced features stay untouched.

I've worked with DTC brands that have been on Klaviyo for three years, paying $400-$700 a month, with zero predictive segments active. The platform has been scoring their customers the whole time. Nobody ever built anything from those scores.

The email revenue gap isn't usually a tool problem. It's a setup problem. The capability is sitting there. Nobody built the segment.

Three things to set up first

If you're on Klaviyo and haven't touched predictive features, here's where to start.

1. Enable smart send time on your existing flows. Go into each active flow and turn on smart sending. Leave it for four weeks before checking open rate data. The model calibrates in 2-4 weeks — looking at week one numbers will give you a false read.

2. Build a predictive churn segment.In Klaviyo Segments, filter by "predicted to churn" + "has placed at least one order." Connect that segment to a 4-email reactivation flow. Make the offer product-specific — not "we miss you," but something connected to what they actually bought. Vague reactivation emails get ignored.

3. Tier your CLV segments differently. Identify your top 20% by predicted lifetime value. Put them in a separate flow with earlier launch access, different incentive thresholds, and higher-value offers. Your highest-CLV customers are worth treating like VIPs. The data to identify them is already in Klaviyo.

This is the part of AI marketing for ecommerce that doesn't show up in product announcements. Klaviyo shipped these features. The gap isn't capability. It's implementation. At Venti Scale, the first thing I do when I take over a Klaviyo account is audit the predictive features — what's active, what's sitting idle, and what the at-risk segment actually looks like right now. That report usually tells the whole retention story.

Dustin Gilmour

Founder of Venti Scale. I've set up Klaviyo for a dozen ecommerce brands. Predictive churn is the first feature I activate after the welcome flow, and the one most accounts have never touched.

About·LinkedIn·X·Updated August 30, 2026