Documentation

Developer Documentation

Welcome to the SaveYourClient Developer Hub. Learn how to connect your data, see which clients are likely to leave, and act first.

Note: You do not need coding skills. Our "Data Studio" handles the mapping for you.

1. How It Works

SaveYourClient uses an "Extract-Load-Predict" pipeline. Your data flows dynamically through our engine to generate actionable insights.

  • 🗄️ Source: Connect PostgreSQL, MongoDB, Oracle, SQL Server, or CSV/SQLite files.
  • ⚙️ Cleaner: Normalizes messy data (dates, currency symbols, missing values).
  • 🧠 AI Model: A Random Forest model trained on your own clients, which scores how likely each one is to leave.
  • 📊 Dashboard: Displays risk scores, financial impact, and actionable retention strategies.

2. Quickstart Guide

  1. Create an account and log in.
  2. Navigate to the Connections tab.
  3. Add your primary data source (PostgreSQL, MongoDB, etc.).
  4. Go to Data Studio and map raw columns (e.g., monthly_fee) to system fields (e.g., MRR).
  5. Click Run Analysis to score your clients.

3. Connecting PostgreSQL

Works with PostgreSQL servers that can be reached from the internet.

postgresql://user:password@host:5432/database_name

Ensure your database allows remote connections from our IP address.

4. Connecting MongoDB (NoSQL)

Nested JSON documents are automatically flattened for easy mapping (e.g., subscription.plan.price).

mongodb+srv://user:password@cluster.mongodb.net/my_db

5. Connecting Oracle Database

Use a DSN or a connection string.

localhost:1521/XEPDB1

6. Connecting SQL Server (MSSQL)

Use the credentials form (Host, Port, User, Pass) for easiest setup.

7. How Scoring Works

Each client gets a Churn Risk Score (0–100%) from a model trained on your own clients. It looks at how much they pay, their plan, how long they have been a client, how recently and how often they were active, and how many support tickets they opened.

Until you have recorded enough real results (around 30 clients you marked Churned or Retained, at least 5 of whom left), the score follows a rule of thumb about inactivity and tickets, and the app tells you so. Open any client to see the factors that pushed the score up or down.

8. Strategy Simulator

Identify at-risk clients and use the Strategy Simulator to test potential interventions.

How it works: Adjust variables like "Logins," "Tickets," or "Days Inactive" to see how the risk score changes.

Example: "If I get this client to log in 3 times this week, will their risk score drop below 50%?"

Why use it: Calculate ROI of retention offers (e.g., is a 20% discount worth it if risk drops by only 1%?).

9. Security & Privacy

  • Encryption: the site is served over HTTPS, and stored database passwords are encrypted.
  • Isolation: every account only sees its own clients.
  • Deletion: you can ask us to delete your data permanently by emailing support@saveyourclient.com.

10. Contact