User retention definitions: 4 analytics tools compared

• Record the cohort membership definition and return method next to every retention chart in an account health review.
• PostHog lets you choose between first time, first ever occurrence, and recurring retention.
• Amplitude offers Return On or After, Return On, and Return On (Custom); Mixpanel retention reports do not extend beyond 60 days.
• Treat documented examples as illustrations, not customer benchmarks, and do not assume equal percentages across tools.

For related reading, see Retention analysis chart: 3 steps across tools and Customer adoption analytics: 4 tools for usage reviews.

The four analytics tools describe different cohort membership choices, return methods, and reporting examples. Amplitude measures returns after an initial event and offers specific return methods like Return On or After. Mixpanel supplies distinct same-week and following-week examples along with a stated reporting limit. Google Analytics describes cohort retention as a time series and uses shared firstSessionDate as a cohort characteristic. PostHog distinguishes recurring, first time within the insight date range, and first ever across all time. Because these definitions vary, record the specific cohort definition and method alongside the retention evidence when preparing an account health review.

When interpreting a chart, identify the selected method and the cohort membership definition used. A Mixpanel chart showing a same-week example is a documented illustration, not a customer benchmark. A Google Analytics cohort defined by firstSessionDate groups users by their initial session date.

Do not assume the tools produce equivalent percentages under their default settings. The example user counts or weights provided by Mixpanel are documented examples, not industry retention benchmarks. User retention is not the same measure as customer renewal or retained ARR. One tool is not the best choice for every customer success team; the right choice depends on the specific review question.

Each analytics platform documents its retention logic differently, requiring you to verify specific settings before interpreting data. Amplitude focuses on the calculation method applied to returns, while Mixpanel highlights a specific temporal boundary for its reports. Google Analytics describes the structural output of cohort data, and PostHog emphasizes the selection of user membership types.

Amplitude allows you to specify the method used to measure retention, offering choices such as Return On or After, Return On, or Return On (Custom) according to Amplitude. Selecting the correct method is essential because it defines how the tool calculates the retention metric for your specific chart configuration.

Mixpanel documentation states that retention reports do not extend beyond 60 days according to Mixpanel. This quote establishes a limit only for Mixpanel; check the reporting limits of any other tool independently.

Google Analytics describes cohort reports as creating a time series of user retention for the cohort according to Google. This structural definition indicates that the output is a sequential dataset tracking retention over time for a defined group of users.

PostHog provides the ability to choose between first time, first ever occurrence, and recurring retention according to PostHog. This selection determines which users are included in the retention calculation, directly impacting the cohort membership and the resulting retention figures.

When preparing an account health review, record which of these specific documentation points applies to the chart you are analyzing. Note whether you are using an Amplitude return method, a Mixpanel 60-day limit, a Google time series structure, or a PostHog membership choice. This ensures the retention evidence aligns with the documented behavior of the specific tool used.

PostHog documentation outlines three distinct ways to define cohort membership for retention analysis, each altering which users appear in the data. The first option is recurring start event membership. According to PostHog, recurring retention counts a user or group as part of a cohort if they performed the start event during that time period, irrespective of if it was their first time or not. This definition includes users who have performed the action previously, as long as they perform it again within the specific cohort period.

The second option is first occurrence within the insight date range. According to PostHog, first time retention only counts a user or group as part of the cohort if it was their first time performing the start event during the insight's date range. Users need not be new to the product; what matters is that this is their first occurrence of the selected start event in the range.

The third option is first occurrence across all history. According to PostHog, first ever occurrence is similar, but only includes users whose very first instance of the start event (across all time, not just the date range) falls within the cohort period.

Google Analytics uses a different grouping characteristic for its cohort specifications. According to Google, users with the same firstSessionDate belong to the same cohort. This groups users based on the date of their first session, rather than the frequency or timing of a specific start event relative to a date range. When interpreting a cohort chart, identify whether the membership rule is based on recurring activity, first activity within a range, or first activity ever, as each definition yields a different set of users. Select the membership definition that matches the specific question you are asking about account health or user behavior.

Amplitude defines the Retention Analysis chart as showing how often users return to your product after they trigger an initial event, according to Amplitude. This method focuses on the timing of the return relative to the first action. When you interpret a chart using this tool, identify the specific return method selected, as the definition hinges on the relationship between the initial event and the subsequent activity.

Mixpanel provides distinct examples to illustrate how buckets function in retention reporting. In the < Week 1 bucket, 6 users did “Sign Up” at any point of the week of Dec 23 - Dec 30, and returned to do “Post Content” in that same week of Dec 23 - Dec 30, according to Mixpanel. This is a same-week return example. In contrast, in the Week 1 bucket, 10 users did “Sign Up” in the same Dec 23 - Dec 30 interval, but returned to do “Post Content” only sometime between Dec 30 - Jan 6, according to Mixpanel. This is a following-week return example. These illustrations clarify that the bucket label depends on when the return occurs relative to the initial cohort period.

The tool also poses specific questions to frame the data, such as how many users come back after month 1 (30 days), according to Mixpanel. This question highlights the time-based nature of the retention query. Mixpanel also separates retained counts from cohort weights. For instance, C_March 1 = 500 (count retained for March 1), according to Mixpanel. Similarly, C_March 2 = 400 (count retained for March 2), according to Mixpanel. These are documented count examples, not customer benchmarks.

W_March 1 = 1000 / 1500 (weight for March 1), according to Mixpanel. W_March 2 = 500 / 1500 (weight for March 2), according to Mixpanel. Treat these as documented cohort weights rather than retained proportions, and do not read them as customer benchmarks.

Amplitude’s approach centers on the method of return measurement, while Mixpanel’s documentation emphasizes the specific bucketing logic and the relationship between counts and weights. Do not treat one publisher's rule as every firm's rule. Record the specific method or bucket definition used in your tool alongside the retention evidence. If your review question concerns the absolute number of returning users, focus on the count examples. If it concerns the share of a cohort that remains active, divide the retained count by its corresponding cohort size.

This example uses round, illustrative numbers and is not tied to any company or product. Suppose a cohort holds 1000 users and 500 of them performed the return event in a later period. The retained count is 500, and the cohort size is 1000, so the retained share is 500 divided by 1000, or one half. If a second cohort holds 1500 users and also shows 500 returning, the count is the same but the retained share is smaller, because the cohort size is larger. Comparing counts alone would hide that difference, so always note the cohort size beside the retained count.

The table below lists the documented retention definitions and examples for the four tools covered in this article. Each row cites the specific source that defines that tool's cohort membership or return method. Use this table to verify which definition applies to the chart you are reviewing.

Record the specific row that matches your current tool to ensure the account health review uses the correct cohort membership logic.

When preparing an account health review, the primary next action is to annotate one retention chart with its specific cohort definition, the selected return method, and any applicable documented reporting limits. This annotation ensures that the evidence presented reflects the exact configuration used to generate the data. For tools like PostHog, the documentation states that users can choose between first time, first ever occurrence, and recurring retention, according to PostHog. Identifying which of these membership choices was selected is critical because each definition captures a different set of users within the cohort.

If the review relies on Amplitude data, the annotation must specify whether the chart uses Return On or After, Return On, or Return On (Custom) methods. Recording the selected method prevents misinterpretation of the retention curve. Similarly, if Mixpanel data is included, the annotation should note the 60-day reporting limit. To connect retention evidence to scoring, see Health score setup: 8 rules for weights and validity.

How do analytics tools define user retention?

Analytics tools define user retention by specifying distinct cohort membership choices and return measurement methods. For instance, PostHog allows users to choose between first time, first ever occurrence, and recurring retention. Meanwhile, Amplitude defines its chart as showing how often users return to a product after triggering an initial event. These structural differences mean that a retention percentage from one tool does not automatically align with a percentage from another without checking the underlying definitions.

What is the difference between first time and first ever retention in PostHog?

First time retention restricts the "first" occurrence to the insight's date range, while first ever looks at all time. First time retention counts a user as part of the cohort only if it was their first time performing the start event during the insight's specific date range, according to PostHog. First ever occurrence includes users whose very first instance of the start event across all time, not just the date range, falls within the cohort period.

Does the stated Mixpanel reporting limit apply to the tool used for my account review?

No, the 60-day limit is documented only for Mixpanel retention reports. The limitation that retention reports do not extend beyond 60 days is specific to Mixpanel, according to Mixpanel. If your account health review relies on a different platform, verify that tool's own reporting limits rather than assuming the 60-day cap applies.

How does Google Analytics group users by firstSessionDate?

Users with the same firstSessionDate belong to the same cohort. This is documented in the time series cohort description, according to Google. The date of the user's first session is the defining characteristic for cohort membership, so the shared first session date identifies which users are grouped together.

Which return measurement methods does Amplitude list?

Amplitude lists three methods: Return On or After, Return On, and Return On (Custom), according to Amplitude. Users specify the method Amplitude uses to measure retention from these options. Identify which of the three is selected when interpreting a retention chart.

This post is general operational information, not legal advice about notice periods or renewal clauses. It does not promise that any method keeps a customer. A review model checked the facts once, and that check can miss errors, so open the sources before you act.