Orders ÷ sessions is the headline, not the whole story
A 3% conversion rate means three orders for every 100 sessions. That is useful for tracking the store overall, but two stores with the same 3% can have very different problems: one loses people before the product page, another loses them at checkout.
The funnel separates those cases. Each step uses the previous step as its denominator, so the rates multiply back to the overall conversion rate: 60% × 30% × 50% × 33.3% ≈ 3.0%.
The four step rates
Session → product view measures whether the landing page and navigation lead people to a product. Product view → add to cart measures product-page persuasion, price and availability. Add to cart → checkout measures cart and shipping-cost friction. Checkout → order measures payment, trust and checkout errors.
A low step rate is not automatically the problem. A step with a naturally small population can still have a large percentage drop; pair the rate with the absolute number of sessions lost.
Worked example
10,000 sessions, 6,000 product views, 1,800 add-to-carts, 900 checkouts and 300 orders. Overall CVR is 3.0%. The step rates are 60.0%, 30.0%, 50.0% and 33.3%. The largest drop-off is product view → add to cart at 70.0%.
That does not mean the other steps are fine. It means the biggest percentage loss is between viewing a product and putting it in the cart, which points to the product page, offer, price or trust signals before checkout.
Two ways to plan
Target CVR mode applies a target rate to current sessions. At 4%, 10,000 sessions should produce 400 orders, a gap of 100 versus the current 300. Required sessions are not meaningful in this mode because the traffic is fixed.
Target orders mode starts from the order goal. For 400 orders at the current 3.0% CVR, required sessions are 400 ÷ 0.03 = 13,333.3. The gap is still 100 orders, but the answer is expressed as traffic needed.
What the model does not do
It does not define sessions for you, split by device or channel, model time lag, handle multiple sessions per order or compare you with an industry benchmark. A session-based funnel also assumes one session produces at most one order.
Use the same analytics definition and period for every input. If your platform counts sessions differently after a tracking change, the funnel will shift even if customer behavior did not.
How to use it
Start with the step that loses the most sessions in absolute terms, then check whether the percentage drop is also large. Test one change at a time and rerun the funnel after enough traffic to see a signal.
For revenue per session and average order value, use the AOV and revenue-per-visitor calculator. This tool deliberately stays on counts and conversion so the two pages do not duplicate each other.