Why this matters
Marketing Analysts use funnels to quantify how people move from awareness to action and where they drop off. Clear funnel visuals help teams prioritize fixes, size opportunities, and measure experiment impact.
- Prioritize where to improve (e.g., checkout vs. sign-up).
- Explain changes to stakeholders without jargon.
- Track experiments step-by-step, not just final conversion.
Progress saving note: You can take the quick test on this page for free. Only logged-in users will have progress saved.
Concept explained simply
A funnel visual shows how many users reach each step of a process and how many drop off between steps.
- Step conversion rate:
next_step_count / current_step_count - Cumulative conversion rate:
final_step_count / first_step_count - Drop-off between steps:
(current_step_count - next_step_count)anddrop_off_rate = 1 - step_conversion
Mental model: "Leaky pipe"
Imagine users flowing through a pipe with leaks at each joint. Each leak is a UX, value, or trust friction. Your job: find the biggest leak that is fixable and impactful.
When to use a funnel vs. alternatives
- Linear journey (A β B β C): Funnel chart.
- Multiple paths or branches: Sankey or flow diagram.
- Time-to-convert patterns: Cohorts or survival curve.
What to show in a funnel chart
- Stage names that map to real actions (e.g., Landing β Product View β Add to Cart β Checkout β Purchase).
- Counts at each step (absolute numbers).
- Step conversion and drop-off percentages.
- Optional: Cumulative conversion to final step.
- Time window and filters (e.g., last 7 days, mobile only).
Minimum viable annotations
- Title with time window and audience (e.g., "Checkout funnel β last 14 days, US only").
- Callouts on the largest drop-off with a brief hypothesis.
- Note on data definition (e.g., "Unique users, first touch in window").
How to build it step-by-step (tool-agnostic)
- Define steps precisely.
- Example: S1 Landing (page_view), S2 Product View (view_item), S3 Add to Cart (add_to_cart), S4 Checkout (begin_checkout), S5 Purchase (purchase)
- Choose a counting unit.
- Unique users per step (most common) or unique sessions
- Set a time window and inclusion rule.
- Example: Users who had a Landing event within the last 14 days
- Aggregate counts by step.
- Compute rates.
- Step conversion:
S2/S1, S3/S2... - Drop-off:
1 - step_conversion - Cumulative conversion:
S5/S1
- Step conversion:
- Pick a visual.
- Bar funnel (horizontal) or waterfall-style for drop-offs
- Annotate the largest leaks and add next actions.
Practical tip: handling repeated users
Use first occurrence per user within the window to avoid double-counting. If you care about repeat journeys, analyze separately.
Practical tip: optional time threshold
Consider a conversion window (e.g., 7 days from first step) so late conversions donβt bias step rates.
Worked examples
Example 1 β Website sign-up funnel
Data (unique users, last 7 days):
Landing 10,000 β Sign-up Start 4,000 β Email Verified 2,800 β Account Created 2,400
- Step conversions: 4,000/10,000 = 40%; 2,800/4,000 = 70%; 2,400/2,800 = 85%
- Largest drop-off: Landing β Sign-up Start (60% drop-off)
- Action idea: Improve CTA clarity on landing; reduce form fields on first screen.
Example 2 β E-commerce checkout
Data (unique users, last 14 days):
Product View 25,000 β Add to Cart 7,500 β Checkout 5,000 β Payment 3,900 β Purchase 3,600
- Step conversions: 30%; 66.7%; 78%; 92.3%
- Largest drop-off: Product View β Add to Cart (70% drop-off)
- Action idea: Add social proof, price clarity, or free shipping threshold messaging on product pages.
Example 3 β Email nurture to trial
Data (recipients in campaign):
Email Sent 50,000 β Open 15,000 β Click 4,500 β Trial Start 1,200
- Step conversions: 30%; 30%; 26.7%
- Largest drop-off: Open β Click (70% drop-off)
- Action idea: Stronger in-email CTA, reduce competing links, align landing page promise.
Choosing the right visual
- Horizontal funnel bars: simplest, good for exec updates.
- Waterfall chart: highlights absolute drop-offs between steps.
- Sankey: when users branch (e.g., different checkout paths). Use sparingly to avoid clutter.
Color and accessibility
- Use one brand color and a lighter shade for drop-offs.
- Always include numbers on bars (counts and %).
- Ensure 4.5:1 contrast for text on bars.
Common mistakes and self-check
- Mixing units (sessions at step 1, users at step 2).
Self-check: Confirm the counting unit is identical across steps. - Missing time window definition.
Self-check: Title includes timeframe and audience. - Only showing percentages without counts.
Self-check: Every label has both. - Hidden sample bias (e.g., mobile-only at some steps).
Self-check: Filters shown and consistent. - Over-precision (e.g., 2.347%).
Self-check: Round to 0β1 decimal.
Exercises (do these, then compare)
These exercises mirror the interactive tasks below. Use any tool (spreadsheet/BI). Keep notes of definitions and filters.
Exercise 1 β Build a sign-up funnel
Dataset (unique users in last 7 days):
- Landing: 8,000
- Sign-up Start: 3,600
- Email Verified: 2,340
- Account Created: 2,145
- Compute step conversions and drop-offs.
- Compute cumulative conversion to Account Created.
- Sketch a horizontal funnel with labels (counts + %).
Exercise 2 β Checkout drop-off diagnosis
Dataset (unique users in last 14 days):
- Add to Cart: 9,200
- Checkout: 5,520
- Payment: 4,000
- Purchase: 3,520
- Find the largest drop-off step and its rate.
- Write two hypotheses explaining that drop-off.
- Propose one quick fix and one experiment.
Checklist before you present
- Time window and audience shown in title.
- Consistent unit (users or sessions) across steps.
- Counts AND percentages on each step.
- Largest drop-off annotated with a next action.
- Data definitions included in a footnote or caption.
Practical projects
- Create a weekly funnel report (one-pager) with a small commentary on what changed and why.
- Run an A/B test on the biggest drop-off step; add a before/after funnel and a short impact summary.
- Segment the funnel by device or channel and show two side-by-side funnels with the same scale.
Who this is for
- Marketing Analysts and growth practitioners needing clear conversion storytelling.
- Product and lifecycle marketers collaborating on experiments.
Prerequisites
- Basic understanding of events and unique users/sessions.
- Comfort with ratios and percentages.
- Ability to aggregate data in a spreadsheet or BI tool.
Learning path
- Define steps and counting logic on a real flow.
- Build a baseline funnel and annotate largest drop-off.
- Segment by device/channel and compare.
- Add a weekly trend of step conversions.
- Design an experiment for the biggest leak.
Mini challenge
Pick a funnel you own. In one slide, show:
- Funnel with counts and step %.
- One segmentation (e.g., mobile vs. desktop) highlighting a difference.
- One prioritized action with expected impact (roughly sized).
Tip: Keep the message to a single, bold callout.
Ready? Take the quick test below. Anyone can take it for free; logged-in users will have their results saved.
Next steps
- Apply these visuals in your weekly business review.
- Pair funnels with qualitative input (session replays or surveys) to explain the "why" behind drop-offs.
- Automate the funnel and annotate changes when experiments ship.