User Behavior Analytics

Historical Context

The funnel is older than software. E. St. Elmo Lewis sketched AIDA — Attention, Interest, Desire, Action — in 1898 as a mental model for advertising. It took over a century for that metaphor to become a literal UI component. Early web analytics tools gave us raw numbers in tables. Then Mixpanel arrived and said: what if we actually showed the drop-off? That tapered shape, wide at the top and narrow at the bottom, became the canonical way product teams visualize intent decaying into action. Amplitude pushed it further. Suddenly you could break funnels by cohort, by device, by the weird edge case your PM swore didn't matter. The visual language solidified: horizontal bars shrinking left-to-right, or stacked trapezoids falling downward. Percentage labels between steps. Red for catastrophic drop-off. The aesthetic is utilitarian — nobody's winning design awards here — but it works because it maps directly to how growth teams think. Every step is a question: why did we lose 40% between signup and activation? Today funnels aren't just retrospective. Teams build them before shipping features, as hypotheses. The diagram becomes the spec. That's the real shift — from reporting tool to thinking tool.

When to Use

Reach for funnel visualizations when you need to expose sequential drop-off in a user journey. They're essential for product analytics dashboards, onboarding flow audits, and any conversion tracking where stakeholders need to see where users bail. Growth teams live in these. Marketing uses them for campaign attribution. UX researchers use them to validate whether a redesigned flow actually reduced friction. If your data isn't sequential or doesn't have a clear start-to-end directionality, pick a different chart — funnels lie when the steps aren't truly linear.

Design Principles

• Make drop-off the loudest signal — the gap between steps should be visually obvious without reading numbers
• Label conversion rates between steps, not just absolute counts — percentages tell the story faster
• Keep step count honest: 3-7 steps max before the visualization loses meaning and becomes decoration
• Use consistent directional flow (top-to-bottom or left-to-right) — never mix orientations within a single view
• Reserve color for anomalies, not decoration — highlight the step that's bleeding users, leave the rest neutral

Technical Specs

Related

Dados & Infografico

Data-Dense Dashboard

Dados & Infografico

Heat Map & Heatmap Style

Dados & Infografico

Executive Dashboard

Dados & Infografico

Real-Time Monitoring

Dados & Infografico

Drill-Down Analytics