What is the right way to calculate anonymous users on my site?

Hi Nitya G., 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> 👋

Short answer: Count unique users who have only anonymous identifiers (no known user_id) within your chosen time window.

How to do this in Mixpanel depends on your ID setup:

- If you use Simplified ID Merge: - Anonymous events have $device_id but no $user_id, and their distinct_id is set to the $device_id (prefixed as $device: in clusters). [Identifying (Simplified); Migrating to Simplified] - Practically, to calculate anonymous users for a period (e.g., MAU): - In an Insights report, choose your “active” event(s), set the measurement to Unique Users (or MAU/WAU/DAU via advanced controls), and filter to events where $user_id does not exist. This yields unique users who were only anonymous during those events in that period. [Insights: Users]

- If you use Original ID Merge: - Anonymous events are tracked under an anonymous distinct_id until you call identify/merge. [Identifying (Original)] - You can similarly count unique users by filtering for events without a known ID (i.e., before $identify is sent), but the docs don’t specify a dedicated filter field beyond distinct_id conventions, so prefer Simplified where possible. The knowledge sources do not explicitly document a ready-made filter for “anonymous only” in Original.

Notes and considerations: - MTU includes anonymous visitors; your MTU count for a month will already reflect unique anonymous and known users combined. To isolate anonymous only, apply the filter above. [MTU billing; MTU overview] - For MAU/WAU/DAU, Mixpanel counts unique users who performed the selected event within the rolling window; use advanced controls in Insights to compute XAU after applying your anonymous filter. [Insights: Users]

If you’re unsure which identity API you’re on, check your project’s ID management version first. [Identifying (Simplified)]

ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.

Let me check the knowledge base for you again. One moment! 📚

You’re on the right track conceptually, but a few Mixpanel identity nuances can make “anonymous users” undercounted if you only filter for userId being null/empty.

What “anonymous” means in Mixpanel - In projects using Simplified ID Merge, anonymous events carry $device_id but no $user_id. When both appear together once, Mixpanel merges them and sets distinct_id to the $user_id going forward. [Identifying (Simplified); Migrating to Simplified] - So “anonymous users” for a time window = unique users who generated events that had $device_id present and no $user_id.

How to calculate in Mixpanel (Simplified ID Merge) - In Insights: 1) Choose the event(s) that define “active” for you. 2) Set the metric to Unique Users (or DAU/WAU/MAU as needed). [MAU in Mixpanel] 3) Add a filter: $user_id does not exist (and optionally $user_id != “” if your pipeline sometimes sends empty strings). This returns unique users who were anonymous when performing those events in that period. [Identifying (Simplified)]

Why your current count can be “surprisingly low” - If users log in during the same period, their events after the login will have $user_id and be counted as identified for those events. Only their pre-login events will meet your “anonymous” filter. This is expected. [Identifying (Simplified)] - If you’re comparing Funnels vs. Insights, remember Funnels can count unique users or conversions, while Insights may be counting total events unless you switch to Unique Users—mismatched aggregations cause discrepancies. [Data discrepancies] - Timezone or filter differences between reports also create gaps—remove filters to baseline, then re-add. [Data discrepancies]

Server-side or Original API notes - Server-side (Simplified): Do not set distinct_id yourself—send $device_id for anonymous events and both $device_id and $user_id after login; Mixpanel derives distinct_id. This ensures proper merging and accurate anonymous filtering by $user_id existence. [Server-side (Simplified)]

- Original API projects behave differently (use $identify/$create_alias/$merge). The sources don’t provide a dedicated “anonymous only” filter beyond identity events, so if you’re on Original, confirm your API version and be cautious interpreting anonymous counts. [Identifying (Original)]

Practical checklist - Verify which ID Merge API your project uses, then apply the method above. [Identifying (Simplified)] - In your events, confirm that anonymous hits truly lack $user_id and include $device_id. - In Insights, switch your metric to Unique Users/MAU before filtering. - If numbers still look off, sanity-check with an unfiltered view, then add the “$user_id does not exist” filter to isolate where the discrepancy appears. [Data discrepancies]

If you share which identity API you’re on and an example event, I can help fine-tune the exact filter.

ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.