I'm curious about the criteria for issuing new sessions in the mix panel. During the session's duration, even if the URL accessed by the user changes and the landing page accessed changes, will the session remain the same, or will it be differentiated and calculated as another session? And does it affect the calculation of the session even if it is moved through various internal links on the website after accessing the web page?
Hi choiyeongung, Please stand by as I search the Mixpanel knowledge base for an AI-powered answer. I normally take about 15 seconds or so. In case AI cannot answer your question, feel free to reach out to Mixpanel support at mixpanel.com/get-support. If you have follow-up questions for this thread, please tag me again with your question: @kapa.ai <follow-up question> 👋
Feel free to ask a follow-up question once I've answered by mentioning me with @kapa.ai
Mixpanel determines when a new session starts based on how sessions are configured in your project. The default and most common method is timeout-based sessions, where a session starts when a user performs any event and ends after a period of inactivity (default: 30 minutes). There are also event-based and property-based session options, but unless you specifically configure those, timeout-based is used.Changing URLs or navigating through internal links within your website does not, by itself, start a new session. As long as the user continues to perform events (such as page views or clicks) within the session timeout window, all those actions—including visiting different pages or landing pages—are considered part of the same session. The session only ends if the user is inactive for longer than the configured timeout period, or if your project uses event-based or property-based session logic that defines session boundaries differently.So, to directly answer your questions: - If the user changes the URL or landing page during a session, the session remains the same (unless the session timeout is reached or a session boundary event/property is triggered). - Navigating through various internal links after accessing the web page does not affect the session calculation—all those actions are grouped into the same session as long as the user remains active within the timeout window.You can configure session settings (timeout duration, event-based, or property-based) in your project settings, and these definitions apply retroactively to your data. For more details, see the official documentation: Sessions: Compute sessions based on existing event data and Session Computation Deep-Dive.
kapa.ai is still learning and improving, please let me know how I did by reacting below.
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
So I'm curious about it more. If the project can set the session boundaries, do you mean that I can configure logic and store the UTM URL inside browser storage to detect the changed event, or save it through the backend through the user's fingerprint and trigger changing session event? I wonder if it's an official event that's basically provided. kapa.ai
Let me check the knowledge base for you again. One moment! 📚
Mixpanel allows you to configure session boundaries in three main ways: timeout-based (default), event-based, and property-based sessions. However, based on the knowledge sources, there is no official feature that automatically triggers a new session when a UTM parameter or URL changes. Here’s what is officially supported: - Timeout-based sessions: Sessions are defined by user inactivity (default: 30 minutes). Changing the URL or UTM parameters does not, by itself, start a new session.
- Event-based sessions: You can define specific events (e.g., "Login" as session start, "Logout" as session end) to mark session boundaries. This is configured in project settings, but it requires you to have clear events that denote session start and end—not based on UTM or URL changes unless you explicitly track such events yourself Sessions: Compute sessions based on existing event data.
- Property-based sessions: You can use a property like session_id to define session boundaries. If you want to start a new session when a UTM parameter or URL changes, you would need to implement logic (client-side or server-side) to generate a new session_id when such a change is detected, and send it with your events. Mixpanel will then treat events with the same session_id as part of the same session, and a new session_id will start a new session Property-Based Sessions.UTM parameters are automatically tracked and stored as event properties and user profile properties by the Mixpanel JavaScript SDK, but they do not, by default, trigger new sessions. You can persist UTM parameters as super properties or disable their persistence for more accurate attribution modeling Track UTM Tags.In summary:
- Mixpanel does not provide an official, built-in event that triggers a new session on UTM or URL change.
- If you want session boundaries to align with UTM or URL changes, you must implement custom logic to generate and assign a new session_id property when such changes occur, and use property-based session computation.
- Storing UTM parameters in browser storage or backend is up to your implementation, not something Mixpanel does automatically for session boundaries.If you need more advanced session logic (e.g., based on UTM changes), you must handle it in your own code and leverage property-based sessions in Mixpanel. This is not an official out-of-the-box event or feature Session Computation Deep-Dive.
kapa.ai is still learning and improving, please let me know how I did by reacting below.
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
I understand that the session is separated after midnight, but if the user was acting between midnight, will this user's session count to 2 sessions when the session data is separated? kapa.ai
Let me check the knowledge base for you again. One moment! 📚
Yes, if a user is active across midnight, Mixpanel will split their activity into two separate sessions. All sessions are reset at midnight according to your project timezone, regardless of the session computation method (timeout-based, event-based, or property-based). This means that if a user is active before and after midnight, their activity will be counted as two sessions—one ending at midnight and a new one starting immediately after midnight.For example, if a user performs an event at 11:50pm and another at 12:05am, the first session will end at midnight, and a new session will start at 12:05am. This behavior is consistent across all session computation types in Mixpanel Session Computation Deep-Dive Examples.
kapa.ai is still learning and improving, please let me know how I did by reacting below.
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
I couldn't write tracking data related to the inflow that needs to be written due to internal reasons. I'm using the growth pricing plan. I want to write data forcibly, but within how many days can I write data? kapa.ai
Let me check the knowledge base for you again. One moment! 📚
On the Growth pricing plan, you can import historical event data into Mixpanel, but there are some important limitations depending on the method you use: - If you use the standard /track API endpoint (including most client-side SDKs), you can only send events with timestamps within the last 5 days. Events older than 5 days will not be ingested via this method Mixpanel SDKs: Node.js Mixpanel SDKs: Ruby Debugging: Delayed Ingestion.- To import events older than 5 days, you must use the /import API endpoint. This endpoint is specifically designed for backfilling historical data and does not have the 5-day restriction. There is no explicit maximum historical window mentioned in the knowledge sources for the Growth plan, but Mixpanel documentation and best practices recommend backfilling up to a year’s worth of data or less for manageability Send Your Data: Backfilling Historical Data.Summary: - /track API: Up to 5 days old - /import API: No strict limit mentioned, but best practice is up to 1 yearIf you need to backfill a large amount of data, consider batching your requests and, if necessary, contact Mixpanel support for higher rate limits during a one-time backfill Quickstart: Track Events.Note: Importing old event data does not incur additional charges if the data is outside your current billing cycle Billing: Old Event Data.
kapa.ai is still learning and improving, please let me know how I did by reacting below.
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
Then, can I use the import api for the growth pricing plan? And I wonder if there are any restrictions on the data that can be written compared to the track api. kapa.ai
Let me check the knowledge base for you again. One moment! 📚
Yes, you can use the /import API on the Growth pricing plan. The /import API is available for all standard Mixpanel plans, including Growth, and is specifically designed for importing historical events older than 5 days, which the /track API does not support Mixpanel SDKs: Node.js Mixpanel SDKs: Android.Restrictions and Differences Compared to /track API:Summary: You can use the /import API on the Growth plan to backfill historical data, and the main differences are in authentication, event age, and request format. The data limits and rate limits are essentially the same as the /track API.If you need to backfill a large amount of data, consider batching and following best practices for rate limiting and error handling Quickstart: Track Events.
kapa.ai is still learning and improving, please let me know how I did by reacting below.
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
Is there any way to define the display name for the data sub-items that are pushed into the list type in the user-defined properties? kapa.ai
Let me check the knowledge base for you again. One moment! 📚
I'm sorry, but according to the knowledge sources, Mixpanel does not support defining display names for the sub-items (elements or keys) within list-type properties in user-defined properties. While you can set display names for top-level event and profile properties using Lexicon, the property names (keys) inside objects or elements within lists cannot have display names assigned in the Mixpanel UI.This limitation is explicitly mentioned for both list of objects and object properties: > Property Names (ie keys) within an object are not supported in Lexicon.
— Supported Properties Data Type: List of Objects> Mainly supported in core reports (i.e. Insights, Funnels, Flows, Retention, Users / Cohorts, Events) as filters and breakdowns. Property Names (ie keys) within an object are not supported in Lexicon. — Supported Properties Data Type: OverviewFor lookup tables, the property names are derived from the CSV header and also cannot be renamed using Lexicon display names Use a Lookup Table.In summary:
You can define display names for top-level properties, but not for sub-items or keys within lists or objects in user-defined properties.
kapa.ai is still learning and improving, please let me know how I did by reacting below.
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
Can I bring the insights items from other projects in my account? kapa.ai
Let me check the knowledge base for you again. One moment! 📚
Yes, you can bring Insights reports (and other reports) from one project to another in your account by using the Move Board feature. Mixpanel allows you to duplicate a Board (which contains your Insights reports) and move it to a different project. This process preserves the reports, filters, and text annotations on the Board. However, before moving, ensure that any custom events, cohorts, or properties used in the reports are also recreated in the destination project, as these do not move automatically with the Board. You may also need to replicate any specific sharing permissions for the moved Board Move Boards: Reuse Boards across different projects Migrating Reports and Non-Data Entities.
kapa.ai is still learning and improving, please let me know how I did by reacting below.
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
I wonder if there is no sharing of users and events between projects at all. kapa.ai
Let me check the knowledge base for you again. One moment! 📚
Mixpanel projects are fully isolated from each other—there is no sharing of users or events between projects. Each project is a separate container for analytics data, and you cannot query or analyze data (users or events) across multiple projects within Mixpanel. All data, including events and user profiles, is stored and managed at the project level, and cross-project analysis is not supported Data Model: How Mixpanel data is organized.If you need to consolidate data from multiple sources, you should send all relevant data to a single project, or consider merging projects if cross-analysis is required Managing Projects.
kapa.ai is still learning and improving, please let me know how I did by reacting below.
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
There are values that are difficult to know from the average value. The deviation between the maximum and minimum values is so severe that I think I need to check the standard deviation, but does the mix panel provide information about this standard deviation? If not, how can I overcome the above situation? kapa.ai
Let me check the knowledge base for you again. One moment! 📚
Mixpanel’s standard Insights reports do not display the standard deviation as a built-in aggregation or measurement option in the UI. The available aggregations for property values include average, median, percentiles (p25, p75, p90, p99), minimum, and maximum, but standard deviation is not listed among them Aggregate Property Additional Aggregation Option Examples.However, if you need to analyze the standard deviation, you can use JQL (JavaScript Query Language), which is Mixpanel’s advanced query tool. JQL provides a built-in reducer called mixpanel.reducer.numeric_summary() that returns count, sum, sum of squares, average, and standard deviation for a collection of numeric values. Here’s an example from the knowledge sources: LBKSfunction main() { return Events({ from_date: "2015-10-01", to_date: "2015-10-02", event_selectors: [{event: "pageview"}], }) .reduce(mixpanel.reducer.numeric_summary('properties.load_time_ms')); }LBKSThis will return an object including the standard deviation for the specified property JQL Builtins.In summary:
- Standard deviation is not available in standard Insights UI reports.
- You can calculate standard deviation using JQL with mixpanel.reducer.numeric_summary().This approach allows you to better understand the spread and variability in your data when averages alone are misleading.
kapa.ai is still learning and improving, please let me know how I did by reacting below.
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.
