Implementing customer health tracking in your Atlassian Marketplace app is essential for understanding how users engage with your product and ensuring they get lasting value. This guide walks you through why and how to implement customer health metrics, the tools you’ll need, practical setup tips, and key privacy considerations.
TLDR - Just give me the recommendations and check out this Video.
Use a Customer Data Platform (CDP) like Segment to centralize and distribute customer health data.
Back-End Setup: Use Segment’s back-end libraries to track key customer behaviors efficiently.
Front-End Setup: Leverage the Segment Analytics library for tracking user interactions while ensuring privacy.
Privacy Best Practices: Remove unnecessary data (like referrer details) to protect user privacy.
Track Key Events: Focus on actions that signal product adoption, value realization, and retention.
Why Track Customer Health Metrics?
Tracking customer health is critical for two key goals:
1. Improving Trial Conversions
If your app has a free trial, tracking engagement helps you understand if trial users are discovering the must-have features that showcase the product’s value. Knowing where users drop off allows you to guide them toward success and increase conversions.
2. Monitoring Customer Health
For paying customers, tracking engagement helps you see who’s thriving, who’s struggling, and who’s at risk of churning. Identifying underused features allows you to proactively offer guidance, training, or support to keep customers engaged.
Choosing the Right Tools: Segment vs. RudderStack
To effectively track customer health, you need a Customer Data Platform (CDP) to collect and distribute data across your tech stack (support, marketing, CRM, etc.).
Why Choose Segment?
Industry Standard: Segment sets the benchmark for data tracking and integration.
Consistency: Ensures your customer health data is structured and reliable across platforms.
Why Choose RudderStack?
Open-Source Flexibility: Allows more customization for teams with strong engineering resources.
Self-Hosting Option: Helps maintain greater control over security and compliance.
Defining and Tracking Key Events
Tracking the right user actions gives you insights into how customers engage with your product. Here’s a breakdown of the three key data calls in Segment and why they matter.
Types of Calls in Segment
1. Identify Call
Links a user to a profile, capturing details like name, email, or account type. Helps personalize the user experience and track engagement trends.
2. Group Call
Links a user to a company or team, allowing you to track product adoption at the account level—essential for B2B SaaS.
3. Track Call
Records user actions, such as completing onboarding steps or activating features. This is the backbone of customer health tracking, showing which users are getting value from your product.
Combining these calls gives a full picture of engagement:
Identify Call → Who is using your product?
Group Call → How is their company/team engaging?
Track Call → What actions are they taking?
What Events Should You Track?
Tracking every action isn’t useful—you need to focus on events that signal engagement and value realization. Here are some key examples:
High-Value Actions (Good Customer Health Signals)
Creating a custom setup: Shows deep engagement (e.g., customizing workflows, setting preferences).
Integrating your product into core processes: Actions like attaching forms to Jira issues indicate long-term adoption.
Editing and refining workflows: Suggests ongoing use and deep product understanding.
Setting up automation rules: Often done by power users—shows strong retention potential.
Low-Value Actions (Not Enough to Measure Engagement Alone)
Logging in: Doesn’t necessarily mean they’re using your product.
Viewing a page: May indicate interest but not active engagement.
Takeaway: Track actions that prove users are gaining value, not just surface-level interactions.
Using Snapshot Statistics for Customer Health Insights
Snapshot statistics provide a clear, high-level view of customer engagement without relying on complex event calculations.
Why Use Snapshot Statistics?
Simple & Reliable: Avoids calculating engagement from multiple event types—just pass a single data point like "total active users this week."
Better Trend Tracking: Monitor week-over-week changes in engagement for early churn signals.
Key Insights to Track
Total number of active users (per account & globally).
Adoption trends over time (e.g., last 7 days vs. previous 7 days).
Changes in feature usage (to detect declining engagement).
Example: Instead of tracking every time a form is created or deleted, just store and send the total number of active forms per account.
Front-End vs. Back-End Implementation
Front-End (Client-Side) Implementation
More transparent (users can see what’s being logged).
Requires privacy safeguards (to avoid capturing sensitive user data).
Best Practice: Remove Unnecessary Data
Use middleware to prevent Segment from collecting sensitive referrer or page title information.
export const segmentClient = AnalyticsBrowser.load( { writeKey: "YOUR KEY" } ); // Remove referrer, search, and title fields from all events segmentClient.addSourceMiddleware(({ payload, next }) => { if (payload.obj.context && payload.obj.context.page) { delete payload.obj.context.page.referrer; delete payload.obj.context.page.search; delete payload.obj.context.page.title; } next(payload); });
Back-End (Server-Side) Implementation
More secure (avoids exposing data in browser traffic).
Might miss user interactions only visible in the front end.
Best Practice: Use a hybrid approach, tracking sensitive data in the back end while capturing user interactions from the front end.
Ethical Considerations: Privacy and Transparency
What Not to Track
User-generated content (e.g., form names). Instead, log anonymized IDs.
Personal data like IP addresses unless explicitly required and compliant with regulations.
Building Customer Trust
Transparency: Publish data tracking details in your Trust Center.
Updated Privacy Policy: Clearly state what’s collected and how it’s used.
Giving Customers Control: Opt-Out Options
Offering an opt-out for analytics is important, but it should be thoughtfully implemented:
Full Product-Level Opt-Out: Allows Jira Admins to disable analytics entirely.
Trial Periods & Free Plans: Consider keeping analytics enabled to help guide users to activation before allowing opt-out.
Example: ProForma (before its Atlassian acquisition) disabled analytics in paid versions but not during free trials—allowing better insights into trial user behavior.
Turning Data Into Action
Once you’ve set up customer health tracking, use the insights to drive real impact:
Spot engagement drops early → Send personalized re-engagement emails.
Identify power users → Offer upsells or invite them to beta programs.
Monitor feature adoption → Improve onboarding for underused features.
Conclusion
Implementing customer health tracking isn’t just about collecting data—it’s about using it to keep customers engaged, reduce churn, and drive product success.
By tracking the right metrics, respecting privacy, and making data actionable, you can build stronger customer relationships and drive long-term growth.
Need an easy way to track and act on customer health data? Accoil makes it simple.
Other reference:
Segment Quick Start Guide for Accoil Analytics