kapa.ai

The knowledge sources don't contain specific information about which Mixpanel plans include the Mirror feature (Warehouse Connectors' Mirror mode). The available documentation about Mirror describes what it does — keeping Mixpanel data in sync with your data warehouse by reflecting additions, updates, and deletions — but doesn't specify plan availability or any recent changes to plan eligibility for this feature. [Mirror intro]

For accurate and up-to-date information on which plans include Mirror, I'd recommend:

Based on the knowledge sources, I can see one likely reason why Mirror sync may not be available in your new project setup.

Mirror Requires Admin or Owner Project Role

To set up Warehouse Connectors (including Mirror syncs), you must have an admin or owner project role on the project. [Warehouse Connectors] If your service account was set up with a lower permission level on the "New Production" project, Mirror mode would not be accessible.

Please verify that the service account (or the user configuring the connector) has admin or owner access on the new project.

Other Things to Check

If none of these apply, the knowledge sources don't contain enough detail to pinpoint the exact cause. I'd recommend reaching out to Mixpanel support directly, as there may be account- or project-level enablement needed on their side.

Based on the knowledge sources available, Mixpanel's case-sensitivity means student and Student are genuinely two separate values — so if your filter for user_type IS student is returning events with Student, that would actually be unexpected behavior.

However, there is one exception worth noting: the == comparison operator in Custom Properties is case-insensitive for strings. [Custom Properties] This means if you're using a Custom Property formula with ==, it would match both student and Student.

For your situation, here's what I'd recommend:

The knowledge sources don't contain enough information to explain why a standard report filter would incorrectly return Student when filtering for student — if this persists after verifying your data, I'd recommend reaching out to Mixpanel support directly.

Sorry Terry Z., I'm currently experiencing some technical difficulties. Please try again in a few minutes. If this issue persists, please reach out to the administrators of this Slack workspace.

The knowledge sources do not contain enough information to answer your question about modifying the end date of an experiment in Mixpanel. What the documentation does mention is that Mixpanel analyzes experiments on the fly and allows you to adjust settings at any time mid-experiment [Experiments setup]. However, there is no specific information about whether you can retroactively set a custom end date for an already-concluded experiment. I'd recommend reaching out to Mixpanel Support directly for clarification on this specific use case.

Based on your previous question and this follow-up, it sounds like you want to count users who did both event A and event B (i.e., an intersection/AND condition).

The best way to do this in Mixpanel is with a Cohort. You can define a cohort in the Cohort Builder with conditions like:

This will give you only the users who performed both events. [Cohorts]

Once you've created that cohort, you can use it as a filter in your Insights report to analyze only those users.

Here's a quick summary of your options:

| Goal | Approach |

| - | - |

| Count A + B (total, may double-count) | Formula A+B in Insights |

| Count users who did A or B (no double-counting) | Custom Event (A OR B) |

| Count users who did both A and B | Cohort with AND condition |

The knowledge sources do not explicitly describe a direct formula approach for an AND intersection count, so using a Cohort filter is the recommended path for your use case.

In Mixpanel's Insights report, when you use a Formula like A+B, it adds the counts of event A and event B together — it does not deduplicate or filter to only users who did both events.

[Insights Advanced]

If your goal is to count users who performed either event A or event B (without double-counting), you should use a Custom Event instead. A Custom Event lets you combine multiple events into a single virtual event, so a user is counted once regardless of which of the combined events they performed. [Custom Events]

In summary:

| Approach | What it does |

|---|---|

| Formula A+B | Adds counts together (can double-count users) |

| Custom Event (A OR B) | Counts unique users who did either event |

The knowledge sources don't contain information about using a "UNION" formula for combining two events in Mixpanel reports or formulas.

What the knowledge sources do cover is the union() method for user/group profile properties. This method appends new values to a list profile property while excluding duplicates, which prevents data duplication in list-type properties.

For example, in JavaScript:

// set profile properties for user "1234"
mixpanel.identify('1234');
mixpanel.people.set({
    name: 'Sam',
    roles: ['sales', 'engineer']
});

// will be ignored since "engineer" already exists in "roles"
mixpanel.people.union('roles', 'engineer');

// add "legal" to "roles" since it doesn't exist yet
mixpanel.people.union('roles', 'legal');

[JS profile updates]

This union() method is available across multiple SDKs (JavaScript, Python, Ruby, Flutter, React Native, Android, Unity, etc.) for both user profiles and group profiles.

If your question is about combining events in a Mixpanel report formula (e.g., in Insights), the knowledge sources do not contain enough information to answer that. I'd recommend checking the Mixpanel documentation directly or reaching out to Mixpanel support for guidance on report-level formulas.

Based on the knowledge sources available, the record_sessions_percent option is described as: "Percentage of SDK initializations that will qualify for replay data capture. A value of '1' = 1%." [Init Options]

The documentation does not explicitly state whether decimal values (e.g., 0.5) are supported for this parameter. Since you've already tested it and found it doesn't appear to work with decimals, this aligns with the fact that the docs only reference whole number examples.

The knowledge sources do not contain enough information to confirm or deny decimal support for record_sessions_percent on the web SDK. I'd recommend reaching out to Mixpanel Support directly to clarify whether sub-1% decimal values are supported, or to request that feature if it isn't.

As a potential workaround in the meantime, you could use the manual capture approach — keeping record_sessions_percent at 0 and programmatically calling mixpanel.start_session_recording() only for a fraction of users based on your own sampling logic in your application code. [Capturing Replays]