Support Documents
Submit a Ticket

Churn Prediction

Suggest Edits

Introduction

Zinrelo’s Churn Prediction Model is an industry-specific solution that analyzes the programs within a given industry to define churn patterns tailored to that sector.

Powered by a sophisticated machine-learning algorithm, this model identifies at-risk loyalty program members, enabling businesses to take proactive measures to retain them.

By analyzing transactional and loyalty data, the model assigns each member a churn risk status, helping businesses segment their audience into high-risk and low-risk members. With these insights, brands can design highly targeted retention campaigns and minimize revenue loss.

Why does churn prediction matter?

The Churn Prediction Model can be strategically leveraged to incentivize both low and high-risk members. By analyzing churn patterns, businesses can craft personalized promotions that retain high risk members and encourage low-risk members to make quicker purchases or spend more.

In summary, churn predictions can also help businesses-

  • Enable personalization- Implement targeted campaigns based on the members' risk profile
  • Boost retention- Engage high-risk members before they stop purchasing.

How does the churn prediction model work?

Data Input

The model is trained on a comprehensive dataset of lifetime transactions and loyalty interactions. This ensures a highly accurate churn prediction mechanism.

It utilizes a combination of:

  • Transactional Data- Includes attributes like total orders, days since last purchase, total revenue, and more.
  • Loyalty Data- Factors in available points, number of redemptions, and overall engagement with the loyalty program.

By analyzing these data points, the model detects behavioral patterns and assigns a churn risk score, helping businesses make data-driven retention decisions.

Data Output

The model assigns churn statuses by analyzing comprehensive dataset:

  • High Risk- Members with a high probability of churning.
  • Low Risk- Members likely to stay engaged.

Applications of churn prediction model

Once the churn prediction model is available for your store, you can leverage it to optimize campaigns and improve member retention. Churn statuses can be found in multiple sections of the Zinrelo admin console, including the Member Page, Activity and Reward Rules, Campaigns, and API & Webhook Responses.

Member Page

You can filter members based on their churn status (e.g., High Risk, Low Risk, Churned). Segment these members, analyze their transaction history, and evaluate their engagement with the loyalty program. You can further use this data to design and run personalized campaigns to re-engage at-risk members.

Activity and Reward Rules

You can define rules that trigger rewards based on a member's churn status. Example: Award 5X points to members with High Risk status and 2X points to members with Low Risk status to encourage engagement.

Campaigns

You can define your campaign’s target audience based on churn statuses. Based on the statuses, run personalized promotions tailored to different risk levels to increase retention.

API & Webhook Responses

The churn status are available in following APIs and webhooks:

API Responses:

Webhook Responses:

Based on the churn_status received in API and webhook response, the dashboard can be dynamically customized for high-risk members—offering targeted incentives, exclusive promotions, or personalized recommendations.

Updated 12 days ago