luvv to helpDiscover the Best Free Online Tools
Topic 2 of 8

Funnel Dashboards

Learn Funnel Dashboards for free with explanations, exercises, and a quick test (for Product Analyst).

Published: December 22, 2025 | Updated: December 22, 2025

Table of Contents

Why this matters

Funnel dashboards show how users progress through key steps (e.g., visit β†’ sign up β†’ activate β†’ purchase). As a Product Analyst, you will use funnels to:

  • Spot bottlenecks and prioritize UX or product fixes.
  • Monitor releases and experiments for step-specific impacts.
  • Forecast outcomes (e.g., signups needed to hit revenue targets).
  • Align teams with a single source of truth for conversion rates.
  • Segment performance by channel, platform, cohort, or geography.

Concept explained simply

A funnel is a sequence of user actions. At each step, some users drop off. Your dashboard counts unique users who reach each step within a defined time window and shows conversion/drop-off between steps.

Mental model: Leaky pipes

Imagine water flowing through connected pipes. Each joint leaks a bit. Your job is to measure leak size at each joint (drop-off), find why it leaks (diagnose), and fix it (experiments). The diameter of the first pipe (traffic) also matters.

Key metrics and formulas

  • Step count: unique users who reached that step.
  • Step conversion rate (n-1 β†’ n): step_n_unique / step_(n-1)_unique.
  • Cumulative conversion rate (1 β†’ n): step_n_unique / step_1_unique.
  • Drop-off rate (n-1 β†’ n): 1 - step conversion rate.
  • Time between steps: median or p90 of time(step_n) - time(step_(n-1)).
  • Re-entry policy: decide whether to count only the first attempt or allow re-entries; be explicit in the dashboard.
  • Attribution window: maximum allowed time from first to last step (e.g., 7 days). Users outside this window are not counted as completed.

Data you need

Your funnel needs user-level events with timestamps.

  • Required fields: user_id (or stable anonymous_id), event_name, event_time (UTC), event_properties (JSON/key-values), platform/channel, session_id (if used).
  • Identity resolution: if both anonymous and logged-in events exist, define a clear merge rule (e.g., stitch anonymous_id to user_id on login).
  • Deduplication: protect against duplicate events (same user_id, event_name, event_time).
  • Time zone: standardize to UTC for storage; present in business timezone on the dashboard.
Edge cases to decide upfront
  • Loops: a user going back to a prior step. Policy: count first valid path per user in the selected time window.
  • Optional steps: mark as non-required; users can skip without being penalized.
  • Parallel paths: ensure step order is logical; enforce ordering by timestamp.
  • Out-of-order events: if ingestion can be late, use an event_time watermark and schedule backfills.

Build a funnel dashboard in 7 steps

  1. Define the question. Example: "Where do users drop between Add to Cart and Payment Success?"
  2. List funnel steps. Name unambiguous events and add filters (e.g., platform = Android).
  3. Choose the attribution window. Common: 24h for checkout, 7–14 days for onboarding.
  4. Compute step entries per user. Use the first timestamp per user for each step within the window.
  5. Calculate metrics. Step counts, step conversion, cumulative conversion, time between steps.
  6. Visualize. Funnel chart + table. Add trend over time and segment filters (channel, platform, country, campaign, cohort).
  7. QA and publish. Compare counts with source events, validate a few user journeys, and document the definitions on the dashboard.
Mini tasks to practice
  • Draft a 4-step onboarding funnel and write plain-language step definitions.
  • Pick a 7-day window and estimate expected completion time distribution (median vs p90).
  • List three segments you will use for breakdowns.

Worked examples

Example 1 β€” Mobile onboarding (4 steps)

Steps: App Open β†’ Sign Up β†’ Email Verify β†’ First Session Complete.

  • Counts: 50,000 β†’ 25,000 β†’ 20,000 β†’ 15,000
  • Step conversion: 50% β†’ 80% β†’ 75%
  • Cumulative conversion to step 4: 15,000 / 50,000 = 30%
  • Median time App Open β†’ Sign Up: 2 min; Sign Up β†’ Verify: 5 min.
  • Insight: biggest drop is App Open β†’ Sign Up (50%). Test shorter form or social login.
Example 2 β€” Ecommerce checkout (5 steps)

Steps: Product View β†’ Add to Cart β†’ Checkout Start β†’ Payment β†’ Order Success.

  • Counts: 200,000 β†’ 40,000 β†’ 32,000 β†’ 25,600 β†’ 24,000
  • Step conversions: 20% β†’ 80% β†’ 80% β†’ 93.75%
  • Segment: Mobile vs Desktop. Mobile Payment β†’ Success is 90% vs Desktop 96%.
  • Insight: Mobile payment friction. Investigate 3DS prompts, wallet options, and errors.
Example 3 β€” B2B activation (6 steps)

Steps: Sign Up β†’ Invite Team β†’ Connect Data Source β†’ Create Dashboard β†’ Share Dashboard β†’ 7-day Active.

  • Counts: 5,000 β†’ 2,500 β†’ 1,750 β†’ 1,050 β†’ 840 β†’ 630
  • Cumulative conversion: 630 / 5,000 = 12.6%
  • Time: Connect Data Source β†’ Create Dashboard median = 45 min (long). Add templates to reduce time-to-value.

Quality checks and diagnostics

  • Sanity: Step 1 count roughly equals unique users triggering that event in the same period.
  • Monotonicity: counts must not increase across steps (unless counting events, which is a red flag).
  • Segmentation stability: totals across segments should sum to overall within tolerance.
  • Latency: if data is late, yesterday’s funnel should stabilize by a known hour.
  • Time math: verify percentiles on time between steps using spot user journeys.
  • Release monitoring: check funnels by app version/build immediately after releases.
QA checklist (tick as you go)
  • I verified unique-user counting per step.
  • I enforced event order and first occurrence per step.
  • I documented the attribution window on the dashboard.
  • I validated 5 real user journeys end-to-end.
  • Segment sums match overall totals within 1%.

Common mistakes (and how to self-check)

  • Counting events, not users. Symptom: counts go up on later steps. Fix: distinct user_id per step.
  • Unclear step definitions. Symptom: arguments about what "Sign Up" means. Fix: add precise event + property filters in the dashboard notes.
  • No time window. Symptom: unrealistic conversions over long periods. Fix: choose a business-relevant window and stick to it.
  • Double counting re-entries. Symptom: conversion > 100% between steps. Fix: use first step occurrence per user within window.
  • Comparing apples to oranges. Symptom: mixing platforms or markets. Fix: segment by platform/country; compare like with like.

Practical projects

  • Project A: Build an onboarding funnel with 4–6 steps, a 14-day window, and device/channel breakdowns. Include median time between steps and a notes panel describing definitions.
  • Project B: Checkout funnel with payment error diagnostics. Add a step-level error rate and a details panel listing top error codes.
  • Project C: Activation funnel with cohort comparison (this month’s signups vs last month). Add trend lines for weekly cumulative conversion.
Minimum acceptance criteria
  • Table: step, users, step conversion %, cumulative %, median time to next step.
  • Charts: funnel snapshot and conversion trend over time.
  • Filters: date range, platform, channel, country.
  • Documentation: step definitions, attribution window, re-entry policy.

Exercises

Complete the exercise below, then use the checklist to confirm quality. Your progress is saved if you are logged in.

  • Exercise 1: Calculate and visualize a 5-step funnel (Visit β†’ Sign Up β†’ Verify β†’ Add Payment β†’ Purchase) with a 7-day window and breakdown by platform.
Practice checklist
  • I computed unique-user counts per step.
  • I added step and cumulative conversion rates.
  • I calculated median time between steps.
  • I added filters for date range and platform.
  • I documented definitions and the 7-day window.

Mini challenge

Your funnel shows a sharp drop from Verify β†’ Add Payment on Android only, starting yesterday. List three immediate checks you will run in the next 30 minutes.

Possible directions
  • Compare by app version; inspect event logs for Add Payment on Android.
  • Review payment SDK errors or 3DS flows on Android.
  • Validate tracking plan: did the event name or property change?

Who this is for

  • Product Analysts and Data Analysts building product performance dashboards.
  • PMs and Designers who need to interpret conversion metrics.
  • Engineers validating event tracking quality.

Prerequisites

  • Basic SQL or comfort with your BI tool’s data model.
  • Understanding of events, users, and unique counting.
  • Access to clean, timestamped product events.

Learning path

  1. Event tracking essentials (events, identities, timestamps).
  2. Define funnel steps and windows.
  3. Compute conversions and time between steps.
  4. Segment and compare cohorts.
  5. Diagnose with QA and monitoring.

Next steps

  • Finish the exercise and confirm the checklist items.
  • Share your dashboard with definitions and agree on a weekly review cadence.
  • Instrument alerts for sudden step-specific changes.

Note: The quick test is available to everyone; only logged-in users have their progress saved.

Quick Test

Ready to check your understanding? Take the short test below.

Practice Exercises

1 exercises to complete

Instructions

Create a funnel for Visit β†’ Sign Up β†’ Verify β†’ Add Payment β†’ Purchase using your BI tool.

  1. Define step events precisely (e.g., Sign Up = event "user_signup"; Verify = event "email_verified"; Add Payment = event "payment_method_added").
  2. Choose a 7-day attribution window starting at the user’s first Visit within the selected date range.
  3. Count unique users per step (first occurrence per user per step within the window).
  4. Compute step conversion %, cumulative %, and median time between consecutive steps.
  5. Add filters for date range, platform (iOS/Android/Web), and acquisition channel.
  6. Create a funnel chart and a table with: Step, Users, Step Conv %, Cumulative %, Median Time to Next.
  7. QA against raw events: ensure monotonic decreasing counts, and verify 5 user journeys manually.
Expected Output
A dashboard page with a funnel visualization and a table showing unique users per step, step and cumulative conversion rates, and median time between steps, plus filters for date range, platform, and channel. Documentation notes include step definitions, a 7-day window, and the re-entry policy (first attempt only).

Funnel Dashboards β€” Quick Test

Test your knowledge with 10 questions. Pass with 70% or higher.

10 questions70% to pass
Product Metrics DashboardsRetention Dashboards

Have questions about Funnel Dashboards?

AI Assistant

Ask questions about this tool