Calculating Average Time Spent on a Page in Mixpanel

Hi event, 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> 👋

You can measure “time on page” in Mixpanel using Sessions and the Session Duration property: Option A: Average session time for a specific page 1) Use the “Session Start” event in Insights, and break down by Current URL (or your page URL property) to isolate the page as a landing page. Then aggregate Session Duration (Seconds) and choose Average. Divide by 60 in a Formula if you want minutes (use the Formula feature: value/60) (Session controls in Insights, Session properties, Insights formulas, Sessions FAQ: average session duration).Option B: Average time spent on a page view event - If you track Page View as a continuous event (or adopt Mixpanel’s automated page view tracking), use the Computed Property “Duration” to measure the time between two occurrences of the same event (works best for Page View within a session). Then, in Insights, select Aggregate Property on the Duration and choose Average; divide by 60 if needed (Computed Properties: Duration, Automated page view tracking reference).Notes - Session Start inherits properties like $current_url automatically, so you can filter/break down by page to get page-level averages (Session properties). - To get “average per user,” switch measurement to “Aggregate Property per User” on Session Duration, then choose Average; use a Formula to convert seconds to minutes (Aggregate Property per User, Sessions FAQ, Measurements).

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! 📚

A common cause is that Session Start only counts sessions whose first event has that page’s URL. If 50+ users viewed the page mid-session (not as the landing page), they won’t appear when you filter Session Start by that URL—so you may see just 1 user whose session actually began there. Session Start is a virtual event that inherits properties (like current URL) from the first event in the session; filtering by URL effectively finds landing pages, not page views (Sessions overview, Session controls in Insights, Sessions use cases: landing page).Alternatives to measure users who interacted with a specific page: - Use your Page View event (e.g., $mp_web_page_view) and filter/break down by Current URL, then measure Unique Users or Total Events in Insights (JS SDK page views, Insights measurements). - In Flows or Funnels, use the page view as a step to capture users who hit that page, choosing Uniques/Totals/Sessions as needed (Flows conversion criteria, Funnels counting methods). - If you specifically want “sessions that contained this page,” in Insights select the page view event and change the aggregation to Sessions; this counts sessions with that event anywhere in the session, not only at start (Session controls in Insights → Aggregation by Sessions).Tip: If you don’t already track page views, enable automatic pageview tracking in the JS SDK or track them server-side to support page-level analysis (JS SDK page views, Server-side page views).

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! 📚

Yes—use a funnel with the page view as Step 1 and the next action as Step 2, then measure Time to Convert between those steps.How to set it up - Build a funnel: Step 1 = Page View (filtered to your specific URL), Step 2 = “Any next event” (or a specific next event if you care about a particular onward action). Use loose/specific order as appropriate. Set a reasonable conversion window (e.g., 30–60 minutes) so late returns don’t inflate times; the window limits how long we’ll wait for the next step to count (defaults to 7 days) (Conversion window). - Measurement: switch to Time to Convert and choose Average/Median/Percentile to get the time users take from Step 1 to Step 2. You can also break down the distribution with the Time to Convert breakdown (Time to Convert, Funnels measurements).Notes and caveats - This measures “time until the next step” within your window; users who don’t hit Step 2 within the window are considered timed out, not part of the conversion time stats (Conversion window behavior). - If “next event” should be anything, consider using Flows to confirm the typical next step from that page, then pin a representative Step 2 in the funnel for cleaner timing (Flows basics). - For a distribution view or to compare segments (e.g., UTM, device), use the Time to Convert breakdown, which lets you filter/bucket times without changing the per-step window (Time to Convert breakdown).

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! 📚

Mixpanel doesn’t provide a built-in “Any next event” step in Funnels. You must specify a concrete Step 2 event. To approximate “any next event,” use one of these approaches: - Use Flows first to discover the most common events immediately after your page view, then pick the top one(s) as Step 2 in your funnel to measure time to convert between Steps 1 and 2 (Flows shows frequent paths to/from events) (Flows overview, Basic features). - Create multiple parallel funnels or a custom event that groups the likely “next events,” then use that custom event as Step 2. The Funnels report supports “Specific order” vs “Any order,” but not “any event” as a step; “Any order” only relaxes sequencing among defined steps, it doesn’t make Step 2 a wildcard (Ordering). For timing, once Step 2 is defined, switch the measurement to Time to Convert and select the step range (e.g., 1→2) and your preferred aggregation (Average/Median/Percentile) (Time to Convert, Measurements).

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! 📚

Yes. In a Funnel where Step 1 is your page view (filtered to that URL) and Step 2 is the next action you care about, you can switch the measurement to Time to Convert and select Average to see the mean time it takes users to go from Step 1 to Step 2. You can also use Median or Percentiles and view a distribution via the Time to Convert breakdown. Be sure to set an appropriate conversion window so only timely next actions are counted (default is 7 days) (Counting methods and measurements, Time to Convert, Conversion window).

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! 📚

- Awareness: Average time to move from Page View (Step 1) to Next Action (Step 2) (use Funnel Time to Convert) (Funnels: Time to Convert, Funnels Quickstart) - Engagement Lag: How long users stay on this page before the next event (Avg Time to Convert 1→2) (Counting methods and measurements) - On-page Dwell to Action: Average seconds from landing on page to subsequent interaction (Time to Convert breakdown)

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! 📚

Yes—build a Specific Order funnel that encodes the exact steps (A → B → C …). In Specific Order, users must complete each step in that sequence to be counted as converted; doing other actions in between is fine, but skipping or reordering required steps will not count as conversion (they drop off at the last completed step) (Ordering: Specific Order, Funnels overview).Tips to tighten validation: - Use exclusion steps to disqualify users who do a “forbidden” action between steps (e.g., exclude Event X between B and C) (Exclusion steps). - Hold a property constant if the same entity must persist across steps (e.g., same page, item, or device) (Hold property constant). - Set an appropriate conversion window so late actions don’t count as valid completions (Conversion window). - Measure conversion rate and Time to Convert to see how many follow the intended flow and how long it takes (Counting methods and measurements).

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! 📚

Yes. Use Custom Buckets to group multiple property values into one segment directly in the Breakdown—no cohorts needed. Add your breakdown, open Customize Buckets, and assign the values you want to merge into the same bucket (e.g., “US” + “Canada” → “North America”). Buckets are non-overlapping and include a “remaining values” catch‑all if desired (Custom Buckets). If you instead need to permanently treat multiple property names/values as one across the project, Owners can merge properties in Lexicon (affects how properties appear in analysis, not raw data) (Lexicon: Merging Properties).

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