What Is a Retention Curve: Complete Definition & Examples

A is a visual chart that tracks what percentage of users from an initial group continue using a product over time. It shows how many people stick around after their first interaction—whether that’s signing up, making a purchase, or taking another key action.

Think of it like tracking a group of friends who join a gym together. The retention curve would show how many are still going to workouts after one week, one month, and three months. Each point on the curve represents a smaller slice of the original group.

Key parts of retention curves:

• Cohort: A group of users who started during the same time window (like all sign-ups from March 1-7)
• Time intervals: The periods you measure (days, weeks, or months after the first action)
• Retention rate: The percentage still active at each checkpoint

The curve starts at 100% on day zero when everyone is new. From there, it typically drops as some people stop using the product. The shape of this decline tells you how sticky your product is.

Customer retention curves reveal patterns that basic miss. While sign-up numbers show how many people try your product, retention curves show how many people actually stick with it.

Product teams use these curves to spot friction points in the . A sharp drop after day three might point to confusing onboarding steps. Marketing teams can compare which bring users who stay longer, not just more signups. Growth teams identify the biggest opportunities to prevent .

The curve also helps . If 40% of users are still active after 30 days, you can estimate how many will remain after 60 or 90 days. This makes planning and budgeting more accurate.

Good retention curves and poor ones have distinct patterns. Most curves fall into four basic shapes, each telling a different story about your product’s health.

Declining curve (trend to zero)

Users abandon the product quickly, and retention drops toward zero over time. This pattern signals fundamental problems—either the product doesn’t deliver expected value, or creates too much friction.

Flat-ish curve

Retention declines gradually but levels off at a low percentage. Some users find ongoing value, but many still churn. This suggests the core product works for a small group, but takes too long for most people.

Flat curve

The ideal scenario in which retention drops initially but then stabilizes at a healthy level. After losing users who weren’t a good fit, the remaining group stays engaged in the long term. This indicates strong product-market fit.

Smile curve

A U-shaped pattern where retention dips and then recovers. This often happens when re-engagement campaigns, new features, or network effects bring users back after initial churn.

Building accurate retention curves requires clean data and clear definitions. The process involves four main steps, each affecting how useful your final chart will be.

Step 1: Define your events

Choose specific actions that represent real engagement, not vanity metrics. “Completed a task” is more meaningful than “opened the app.” Pick one action as your starting point (like account creation) and another as your return signal (like logging in or using a core feature).

Keep these definitions consistent across all cohorts you want to compare. Changing what counts as “active” halfway through makes your data unreliable.

Step 2: Select your cohort

Group users who started at the same time—weekly cohorts work well for most products. You can segment by acquisition source, device type, or user characteristics to understand what drives different retention patterns.

Remove obvious outliers like test accounts or bots that might skew your results.

Step 3: Choose your time intervals

Match your measurement periods to how often people naturally use your product. Daily intervals work for social apps or games. Weekly or monthly intervals make more sense for productivity tools or ecommerce sites.

Step 4: Plot and analyze

Use analytics tools to visualize your data. Amplitude Analytics lets you build retention curves by selecting your start and return events, setting time intervals, and comparing different cohorts side by side.

Look for the steepest —these show where you’re losing the most users and where improvements could have the biggest impact.

The shape of your curve points to specific problems and opportunities. Focus on three key areas when reading your retention data.

Find the biggest drops: The steepest decline usually happens in the first few days or weeks. This is where most users decide whether your product is worth their time. Mark these inflection points and investigate what happens right before them.

Measure time-to-value: Compare users who reach key milestones early (like completing their profile or using a core feature) with those who don’t. Users who hit these markers faster typically show better long-term retention.

Separate early and late retention: Day 1 and week 1 retention reflect onboarding effectiveness. Month 3 and beyond show whether people form lasting habits with your product.

To understand why retention drops, look beyond the curve itself. Use to see the most common paths before churn. Session replay shows exactly where users get stuck or confused. Unlike point solutions that only show you the numbers, comprehensive platforms connect retention data with user behavior insights.

Once you’ve identified drop-off points, you can test specific changes to flatten the curve. These strategies target the most common reasons people stop using products.

1. Streamline onboarding

Remove unnecessary steps that delay the first moment of value. Each extra form field or required action increases the chance someone will quit before experiencing what makes your product useful.

Common improvements:

• Reduce sign-up fields to the minimum needed
• Defer optional setup steps until later
• Show progress indicators during multi-step processes

2. Highlight valuable features early

Guide new users toward actions that correlate with long-term retention. If people who follow five accounts in their first week stick around longer, make following accounts easier and more prominent.

Use progressive disclosure to introduce complexity gradually rather than overwhelming new users with every feature at once.

3. Add contextual help

Trigger when users show signs of confusion—like repeated clicks on the same area or long pauses on key pages. Contextual tooltips work better than generic welcome tours that most people skip.

4. Test changes with controlled experiments

Use feature flags to roll out onboarding improvements gradually. Compare retention between the new experience and the old one to confirm your changes actually help.

prevents you from making changes that feel logical but don’t actually improve retention rates.

5. Re-engage churned users

Target people who were previously active but haven’t returned recently. that remind users of saved progress or new features relevant to their past behavior can bring people back.

Time these messages based on your retention curve—if most people churn after two weeks of inactivity, reach out after 10 days.

Retention curve analysis works best when connected to other user behavior data in a single platform. Amplitude's Digital Analytics Platform combines retention analysis with journey mapping, session replay, and experimentation tools.

Instead of using separate point solutions for each type of analysis, you can build retention curves, identify drop-off points, watch user sessions, and test improvements all in one workflow. This eliminates the complexity of stitching together multiple tools and gives you faster insights.