This article outlines how filters work within Funnels—what they apply to, how they affect the data, and the logic behind how filtered audiences are calculated.
Use this guide to understand exactly how filters like Device, Referrer Group, or Date impact your step counts, visitor flows, and conversion rates—so your funnel insights are accurate, intentional, and aligned with your analysis goals.
Leaderboard filters apply to all funnels visible in the table. For example, if you use the checkboxes to select 5 funnels to compare, then click ‘View Selected Only’, you could view 5 funnels and only apply filters to those funnels.
In the funnels Dashboard, you are always viewing one funnel only. The filters will only apply to that single funnel.
Definition: Filters to only include visitors where at least the Step 1 event occurred within the selected date range.
Calculation: Accept event E if window_start ≤ E.timestamp ≤ window_end.
Impact on Total Population: Removes any user where step 1 event is outside the date range. Step 1 count (denominator) only includes users whose Step 1 falls within this window.
Example: If Funnel Date Range = Mar 1–Mar 31, a user who started Feb 28 or finished Apr 1 is excluded. If a user did step 1 on March 15 and completed the funnel on March 30, they would be fully included in this date range. If a user did step 1 on March 15 and completed the other steps in April, only their data in March would be included in this date range.
Notes:
If selected first, First Step Date Range must start on the same day as Funnel Date Range.
Last Step Date Range must end on the same day as Funnel Date Range.
Cannot select "today."
Marketer’s Question: "Show me funnel data that happened this month only."
Definition: Filters users whose Step 1 occurred within a specific time window.
Calculation: Compare Step 1 timestamp to the selected range.
Impact on Total Population: Excludes any user whose Step 1 falls outside the range. Total funnel entries are reduced accordingly.
Example: Step 1 between Mar 5–10; user who started Mar 3 is excluded.
Notes:
If selected first, Funnel Date Range must start on the same day.
Last Step Date Range must not exceed this window.
Cannot select "today."
Marketer’s Question: "How did cohorts that entered during launch week perform?"
Definition: Includes only users whose final funnel step occurred in a selected date range.
Calculation: Compare timestamp of final step (Step N) to the selected range.
Impact on Total Population: Does not change the denominator, but the number of visitors that complete (numerator) is reduced if users finish outside the window.
Example: Only count final completions that occurred Mar 20–31.
Notes:
If selected first, Funnel Date Range must end on the same day.
First Step Date Range must not exceed the Last Step range.
Cannot select "today."
Marketer’s Question: "Who actually purchased before quarter-end?"
Definition: Filters to users whose Step 1 session referrer matches the selected referrer.
Calculation: Compare Session Referrer for Step 1 to chosen referrer group.
Impact on Total Population: Narrows the starting audience to visitors whose Step 1 session referrer = X. Non-matching users are removed entirely.
Example: 2,500 of 10,000 starters have referrer = “Social”.
Marketer’s Question: "How do paid social entrants convert versus other traffic?"
Definition: Filters to users whose last step of the funnel matches the selected referrer.
Calculation: Compare Session Referrer for final step to chosen referrer group.
Impact on Total Population: Denominator remains unchanged; completion count is filtered by referrer.
Example: 300 of 2,000 starters finish in “Organic Search” sessions.
Marketer’s Question: "How often do buyers finish this Funnel via organic search?"
Definition: Includes users who triggered at least one funnel step on a desktop device.
Calculation: Include if any funnel step occurred on a desktop.
Impact on Total Population: Narrows funnel to only those who used desktop; others are excluded from funnel analysis.
Example: Selecting "Mobile" retains 6,500 of 10,000 starters.
Marketer’s Question: "Does the funnel work better on mobile or desktop?"
Definition: Includes users who triggered at least one funnel step on a mobile device.
Calculation: Include if any funnel step occurred on a mobile device.
Impact on Total Population: Narrows to mobile users only.
Example: Selecting "Desktop" retains 3,500 of 10,000 starters.
Marketer’s Question: "Does the funnel work better on mobile or desktop?"
Definition: Includes only users who complete all funnel steps in a single session.
Calculation: Group by session_id; require all steps in same session.
Impact on Total Population: Total starts are not impacted, but subsequent steps are only counted if they occur in the same session as step 1
Example: 10,000 start; 3,200 complete in same session → CR = 32%.
Marketer’s Question: "What is the one-session conversion rate?"
Definition: Includes only users whose last funnel step occurs in a different session from Step 1.
Calculation: Keep if session(final_step) ≠ session(first_step).
Impact on Total Population: Total starts are not impacted, but subsequent steps are counted if the visitor’s last step (not necessarily final step) occurs in a different session than the first step).
Example: Step 1–2 in Session 1, Step 3 in Session 2; user is counted. 10,000 start; 1,100 finish in different session → CR = 11%.
Marketer’s Question: "How many users need multiple visits to convert?"
Definition: Includes users whose Step 1 occurred in their very first session ever recorded.
Calculation: first_session_time == session(Step 1)
Impact on Total Population: Narrows to users who started the funnel as true new users. Denominator = Only those whose earliest recorded session includes Step 1. Excludes anyone with a prior session.
Example: 6,100 new visitors remain.
Edge Cases: Requires accurate session history. Anonymous users may be undercounted.
Marketer’s Question: "What’s the funnel for new users only?"
Definition: Includes users who had at least one session before Step 1.
Calculation: sessions_before_step1 ≥ 1
Impact on Total Population: Narrows funnel to returning users only. Denominator = Only those who had a session before Step 1. Excludes brand-new visitors whose Step 1 was their first session.
Example: 3,900 returning visitors remain.
Edge Cases: Requires accurate session history. Anonymous users may be undercounted.
Marketer’s Question: "Do returning users convert better?"
Definition: Only includes users who had Step 1 as their very first recorded event.
Calculation: first_event_time = step1_time
Impact on Total Population: Narrows funnel to first-time visitors only. Denominator = only those who have Step 1 as their earliest tracked event.
Example: 6,100 of 10,000 starters are first-time visitors.
Marketer’s Question: "How do brand-new visitors convert vs returning?"
Definition: Allows Step 1 to be any event in a visitor’s history.
Calculation: No special check / extra rule.
Impact on Total Population: Includes all visitors who started the funnel.
Example: Default setting; broadest inclusion.
Marketer’s Question: "Include all visitors, even if they’ve been here before."
