A conversion funnel is a model of the sequence of steps a visitor takes toward a desired action, usually a purchase, together with the number of people who reach each step. In e-commerce it typically runs from landing on the site through product views, adding to cart and checkout to a completed order, and it shows where shoppers drop out along the way.
The funnel shape reflects a simple fact: every step loses some visitors, so the group that reaches the end is much smaller than the group that started. In conversion work, the funnel turns this loss into measurable step-by-step rates that point teams toward the stages where improvement will have the greatest effect on revenue.
How a conversion funnel is built
A conversion funnel is built by defining ordered steps as tracked events and counting how many sessions or users complete each one within a chosen time window. Analytics tools then calculate the step conversion rate, the share of people who move from one step to the next, and the overall conversion rate from the first step to the last.
Definitions matter. A funnel can count sessions or users, can require steps in strict order or allow skipped steps, and can use windows ranging from one session to several weeks. Changing any of these choices changes the numbers, so a funnel is only comparable over time when its definition stays fixed.
| Stage | Tracked event | Main question it answers |
|---|---|---|
| Visit | Session start or landing page view | Does traffic arrive at all |
| Product view | Product page view | Do visitors find relevant products |
| Add to cart | Cart addition event | Does the product page convince |
| Checkout start | Checkout page view | Does the cart lead to commitment |
| Payment details | Shipping and payment step completed | Do costs and forms cause friction |
| Purchase | Order confirmation | Is the transaction completed successfully |
Reading funnel metrics
The most useful reading compares step conversion rates with each other and with earlier periods, rather than focusing on the overall rate alone. Overall store conversion rates commonly fall somewhere around one to four percent of sessions, but they vary strongly by product category, price level, device and traffic source, so external benchmarks are only a rough guide.
| Metric | Calculation | What a weak value suggests |
|---|---|---|
| Overall conversion rate | Purchases divided by visits | General mismatch of traffic, offer or experience |
| Product view rate | Product views divided by visits | Poor navigation, search or landing page relevance |
| Add-to-cart rate | Cart additions divided by product views | Weak product pages, price or availability issues |
| Cart abandonment rate | Carts without purchase divided by carts | Unexpected costs, forced account creation |
| Checkout completion rate | Purchases divided by checkout starts | Form friction, payment failures, trust concerns |
Because the steps multiply, a small gain early in the funnel can matter as much as a large gain late in it. Raising a 50 percent step rate to 55 percent increases final orders by ten percent, assuming the later steps stay constant, which helps teams compare opportunities in a common unit.
Conversion funnel compared with related models
The conversion funnel is a measurement model for on-site behavior, which distinguishes it from broader marketing frameworks. Marketing funnels such as awareness, consideration and decision describe the whole buying process across channels, while the e-commerce conversion funnel measures concrete events inside the store.
| Concept | Focus | Typical data |
|---|---|---|
| Conversion funnel | Step-by-step progress toward purchase | Event counts and step rates |
| Marketing funnel | Stages of awareness and intent across channels | Reach, engagement, leads |
| Customer journey map | Experiences, emotions and touchpoints | Research and qualitative insight |
| Micro-conversion | Smaller actions inside or beside the funnel | Newsletter sign-ups, wishlist additions |
| Conversion rate optimization | Systematic improvement of conversion | Experiments, session recordings, tests |
Real shoppers rarely move in a straight line. They leave, compare prices, return days later on another device and buy. A funnel simplifies this into a linear model, which is useful for diagnosis but should be complemented by path analysis and cohort views that show non-linear behavior.
Using the funnel to improve conversion
Teams use the funnel to locate the largest leak, form a hypothesis and test a fix. The funnel shows where people leave; session recordings, surveys and usability tests explain why; controlled experiments confirm whether a change actually helps.
| Weak stage | Frequent cause | Typical remedy |
|---|---|---|
| Visit to product view | Irrelevant landing pages or poor search | Better landing page match and site search |
| Product view to cart | Missing information, weak images, unclear price | Richer product content, visible delivery terms |
| Cart to checkout | Surprise shipping costs | Show costs early, free shipping threshold |
| Checkout to purchase | Long forms, forced registration | Guest checkout, digital wallets, fewer fields |
Segmenting the funnel by device, traffic source and new versus returning visitors often reveals that a problem affects only one group, for example a payment step that fails on certain mobile browsers. An unsegmented funnel would hide such issues inside an average.
Common mistakes in funnel analysis
Funnel analysis is only as reliable as the tracking and definitions behind it. Many misleading conclusions come from broken events or from comparing funnels built in different ways.
- Broken or duplicated tracking events that create impossible step rates above 100 percent.
- Ignoring consent-related data gaps, which can undercount visits and distort rates.
- Optimizing one step in isolation, such as pushing more carts that never convert into orders.
- Treating correlation as cause without testing whether a change really drove the improvement.
The simple explanation
A conversion funnel shows how many shoppers make it through each step of buying. Picture a wide funnel: many people come in at the top by visiting the shop, fewer look at a product, fewer still put it in the cart, and only some finish paying.
By counting people at each step, a shop can see exactly where most of them give up. That is the step worth fixing first, much like checking which part of a leaky pipe is losing the most water.
How it is used in practice
An online store selling kitchen appliances reviews its monthly funnel. Of 200,000 sessions, 90,000 reach a product page, 9,000 add to cart, 5,400 start checkout and 2,700 complete a purchase. The overall conversion rate is 1.35 percent, and the checkout completion rate of 50 percent stands out as unusually low.
Segmenting by device shows that the checkout drop is concentrated on mobile. Session recordings reveal that a mandatory account registration step and a long address form cause many mobile users to quit. The team introduces guest checkout and an express wallet payment option and tests it against the old flow for four weeks.
In the test, mobile checkout completion rises noticeably while desktop remains stable. Because the later funnel steps did not change, the gain flows directly into additional orders, and the new checkout becomes the default.