What is a conversion funnel? Ecommerce funnel explained
A conversion funnel is the sequence of steps a shopper takes from first visit to purchase, for example product view, add to cart, begin checkout and order. A funnel report shows how many people reach each step, so you can see where most of them drop out.
A typical ecommerce funnel
If your store sends Google Analytics 4's recommended ecommerce events, the steps map naturally to them:
- Product viewed (view_item)
- Added to cart (add_to_cart)
- Checkout started (begin_checkout)
- Order placed (purchase)
Hypothetical example:
| Step | Users | Continue to next step |
|---|---|---|
| Viewed a product | 10,000 | 10% |
| Added to cart | 1,000 | 50% |
| Started checkout | 500 | 60% |
| Purchased | 300 | – |
End to end, 300 of 10,000 product viewers bought, a 3% rate. The largest drop is between viewing and adding to cart, which points to the product page (price, photos, shipping information) as the first place to look. The 40% who start checkout but do not finish are part of your cart abandonment.
Open and closed funnels
GA4's funnel exploration lets you choose. In a closed funnel users must enter at the first step and are only counted in the steps they complete in order. In an open funnel users can enter at any step, which suits stores where many people land directly on a product or cart page. The same data can look quite different under each setting, so note which one a report uses. GA4 also offers a trended view to see how each step changes over time.
Why it matters
Your overall conversion rate is the product of every step's rate. Improving the weakest step usually moves the total most, and a funnel tells you which step that is instead of leaving you to guess.
Common mistakes
- Tracking gaps: if one step's event does not fire on some devices, the drop you see is a measurement bug, not shopper behavior.
- Too many steps: long funnels get thin data at the bottom; start with four or five.
- One funnel for everyone: break it down by device and traffic source, since mobile and desktop shoppers often stall at different steps.