Funnel Analysis

Funnel analysis measures how users progress through a defined sequence of steps, tracking the conversion rate and drop-off at each stage. The classic example is an e-commerce purchase funnel: Product Page Viewed > Added to Cart > Checkout Started > Payment Entered > Order Completed. If 10,000 users view a product, 3,000 add to cart, 1,500 start checkout, 1,200 enter payment, and 1,000 complete the order, you now know exactly where the biggest leaks are.

Why it matters: funnels reveal the highest-leverage opportunities for improvement. In the example above, the biggest drop-off is from viewing to adding to cart (70% drop). Improving that single step by even 10% has a cascading effect on all downstream conversion. This is far more efficient than trying to improve every step equally. Funnels also let you quantify the revenue impact of each drop-off point, which helps prioritize engineering and design resources.

How to build funnels: define the steps as events (using your event tracking implementation). In tools like Mixpanel, Amplitude, or Kissmetrics, you select the events in order, set a conversion window (the maximum time allowed between first and last step), and the tool calculates conversion rates. You can then segment funnels by user properties (device type, acquisition source, plan tier) to find patterns.

Advanced techniques: compare funnels across segments to find where specific user groups struggle. Look at time-to-convert between steps, not just whether users convert. A user who takes 15 minutes between steps 2 and 3 might be confused, while one who takes 30 seconds is flowing smoothly. Also build "negative funnels" that track paths leading to unwanted outcomes (cancellation, refund request, support ticket).

Common mistakes: defining funnels that are too long, which makes it hard to diagnose issues. Not setting appropriate conversion windows: a 30-day window for an e-commerce checkout is too loose. Only looking at the overall funnel without segmenting. Treating the funnel as purely linear when users often loop back, skip steps, or take detours.

Practical example: a SaaS company builds an onboarding funnel: Account Created > Profile Completed > First Project Created > Teammate Invited > First Report Generated. They discover that only 22% of users who create a project go on to invite a teammate. They add an in-app prompt after project creation that says "Share this project with your team" and the invite rate jumps to 41%, directly improving downstream retention.