Funnel Analysis
Also: Conversion Funnel Analysis, Sales Funnel Analysis, Marketing Funnel, Funnel Metrics
Funnel analysis tracks how users move through a sequence of steps toward a goal, revealing where they drop off and which stages need improvement.
What It Is
Funnel analysis is a method for measuring how users progress through a defined series of steps toward a desired outcome, such as a purchase, signup, or subscription. The term "funnel" reflects the natural narrowing that occurs as people move from a broad starting point (for example, all website visitors) toward a smaller set of users who complete the final action.
Each step in the funnel records the count of users who reached it, allowing you to calculate conversion rates between consecutive steps and an overall conversion rate from start to finish. The largest gaps between steps signal where users abandon the process.
Why it matters
Funnel analysis turns a vague sense that "conversions are low" into specific, actionable insight about where and how much value leaks out of a process. It matters because:
- It pinpoints the highest-impact areas for optimization rather than spreading effort thinly.
- It quantifies the revenue or growth opportunity tied to fixing a given step.
- It creates a shared language across marketing, product, and finance teams.
- It supports forecasting, since stable step conversion rates help predict outcomes from top-of-funnel volume.
How it is used in practice
1. Define the steps. Choose an ordered sequence of events (for example: visit, view product, add to cart, checkout, purchase).
2. Set the scope. Decide on the time window, the user population, and whether steps must occur in strict order.
3. Measure conversion and drop-off. Calculate the percentage advancing at each step.
4. Segment. Break the funnel down by channel, device, geography, or cohort to find where performance differs.
5. Diagnose and act. Form hypotheses about why a step underperforms, run experiments (such as A/B tests), and re-measure.
Funnels can be open (users may skip steps) or closed (strict order required), and time-bound (steps must occur within a set period).
Concrete Example
An e-commerce team observes:
- Visited site: 100,000
- Viewed product: 40,000 (40%)
- Added to cart: 12,000 (30%)
- Started checkout: 6,000 (50%)
- Completed purchase: 3,000 (50%)
The overall conversion rate is 3%. The biggest relative drop is from view to cart (only 30% advance). If the team improves that single step to 40%, completed purchases rise to roughly 4,000, a 33% lift, without increasing traffic. This focus on the weakest stage is the core value of funnel analysis.
Frequently asked questions
What is funnel analysis?
Funnel analysis is a method for measuring how users progress through an ordered series of steps toward an outcome such as a purchase, signup, or subscription. Each step records how many users reached it, so you can compute conversion rates between consecutive steps and an overall rate from start to finish. The largest gaps between steps show where users abandon the process.
Who uses funnel analysis inside a company?
Marketing, product, data, and finance teams all use it, which is part of its value: it creates a shared language about where value leaks. Marketing looks at channel and top-of-funnel volume, product looks at the steps users abandon, finance uses stable step conversion rates for forecasting. Funnel analysis is a core topic for CMO and CDO-type roles that own conversion and measurement.
What is the difference between an open funnel and a closed funnel?
In a closed funnel, users must complete the steps in strict order to be counted; in an open funnel, they may skip steps and still appear at a later stage. Funnels can also be time-bound, meaning all steps must occur within a defined period. The choice changes your numbers, so it should be fixed before you compare results across periods or channels.
Where do I start if my conversion rate is low and I don't know why?
Start by defining an ordered sequence of events, then set the scope: time window, user population, and whether the steps must occur in strict order. Measure the percentage advancing at each step, then segment by channel, device, geography, or cohort to see where performance differs. Only then form hypotheses about the weakest step and test them, for example with A/B tests, before re-measuring.
Is it better to fix a weak funnel step or to buy more traffic?
Fixing the weakest step is often cheaper, and the arithmetic shows why. In a typical e-commerce funnel of 100,000 visits, 40,000 product views, 12,000 add-to-carts, 6,000 checkouts started and 3,000 purchases, the overall conversion rate is 3% and the worst relative drop is view-to-cart, where only 30% advance. Raising that single step from 30% to 40% lifts purchases from 3,000 to roughly 4,000, a 33% gain with no additional traffic.