Setup Guide

From raw behavioral events to daily churn and LTV predictions in four steps.

No data team, no ML engineering, no months of setup. If you have a Shopify store, a Stripe subscription feed, or a web app with a login event, you can have live propensity scores within a week. Here is exactly what that looks like.

Day 1 Connect your data source (~30 min)
Day 2-7 Segmentvue ingests behavioral history
Day 7+ First predictions available
Day 14+ Model improves with more events
1

Connect your event data

Three ingestion paths. Pick the one that fits your stack today. You can add a second source later. The events that matter most for churn prediction are: logins, page depth, order completions, plan changes, email opens, and cancellation attempts. If your stack generates those, you have what the model needs.

snippet.js
// Paste before </head>
<script>
  !function(s,v,u,e){
    s['Svue']=s['Svue']||{};
    s.Svue.load('wk_YOUR_KEY');
  }(window);
</script>
SQL sync (Postgres, Bigquery, Snowflake)
Webhook ingest (Stripe, Shopify)
2

Define your segments

The rule builder works against any attribute in your unified customer profile: recency, frequency, monetary value, email engagement rate, feature adoption depth. You can define a segment in a few minutes without writing SQL. Once defined, the segment population updates continuously as new events arrive. It is never a static snapshot.

Before

1 segment: Active Customers

After

5 segments: High-Value Actives, Price-Sensitive, Lapsing, Recently Churned, Recovering

3

Get propensity scores and LTV forecasts

After 7 days of ingestion, Segmentvue trains the initial churn propensity model and 12-month LTV forecast for your tenant using your own behavioral history, not an industry average. Every customer profile receives a daily-updated score. The model continues improving as it observes more events: confidence bands tighten noticeably by day 30.

73%
Churn Probability Updated each day as new events arrive. Score above 65% triggers activation rules you define.
$1,840
Predicted 12-mo LTV 12-month revenue projection per customer based on order frequency, AOV trend, and plan trajectory.
4

Sync scores to your activation stack

Segmentvue does not replace your email tool or CRM. It feeds them. Configure a sync rule: when a customer's churn propensity crosses a threshold, push them to a win-back list in your ESP, write the score to a CRM contact field, or fire a webhook to any downstream system. Your lifecycle automation then runs exactly as you designed it, but on daily-fresh data instead of last month's export.

At-Risk Segment
Churn prob: 65%+
Win-back sequence fires
Daily scores
CRM sync
Sales contact enriched

A realistic setup timeline

We do not promise a churn model on day one. Here is what actually happens, and when.

Day 1 Install the JS SDK or configure SQL sync to your warehouse. Most teams complete this in about 30 minutes. Webhook ingest from Shopify or Stripe takes about the same. Events start flowing immediately.
Day 2-7 Segmentvue backfills your behavioral history. All plans support up to 24 months of historical events. This backfill is what gives the model enough signal to produce accurate propensity scores from day one.
Day 7+ First churn propensity scores and 12-month LTV forecasts calculated for every customer profile. Segment populations populated. Sync rules can be activated.
Day 14+ The model has now observed two weeks of live behavioral events on top of the historical backfill. Confidence bands narrow. Score volatility drops. Segment membership stabilizes as the model learns your customers' normal patterns.
Start free trial See pricing