Ecommerce Conversion Funnel Analytics: How to Find Conversion Drop-Offs

An ecommerce conversion funnel is the path from arriving at a store to buying, and buying again. Its stages are acquisition, landing, discovery, product view, add to cart, cart, checkout, payment, purchase and retention. To find where customers drop off, track an event for each stage and calculate the rate from each stage to the next. Segment those rates by device, traffic source, landing page and new versus returning customers. The stage and segment with the largest avoidable drop is where to investigate first, using recordings, heatmaps and user testing to learn why.

The diagram above shows the stages. Each needs a measurable event and a stage-to-stage rate.

GA4's recommended ecommerce events cover the journey from view_item_list to purchase and refund. Many platforms and integrations send them automatically, but check that each fires once, with the right items and values. Shopify's own conversion rate breakdown shows a four-step version: sessions, sessions with cart additions, sessions that reached checkout and sessions that completed checkout.

In GA4, funnel explorations let you build the funnel from these events. A closed funnel counts only users who start at the first step; an open funnel lets users enter at any step, which suits stores where many shoppers land directly on product pages.

The overall conversion rate compresses the whole funnel into one number. Stage-to-stage rates are more useful: product views per session, add-to-carts per product view, checkouts per cart, purchases per checkout. A falling overall rate with a stable add-to-cart rate and a falling checkout completion rate points straight at checkout.

The following is a hypothetical example to show the method; the numbers are illustrative, not benchmarks. A store sees its overall conversion fall over two months. Its funnel, split by device, looks like this:

The drop is isolated to mobile, between adding to cart and starting checkout. That rules out traffic quality, product pages and checkout itself as the main cause and points to the mobile cart. The next steps are to watch mobile recordings of shoppers who added to cart but didn't start checkout, check recent changes to the cart drawer or apps, test on phones and look for errors. In this hypothetical case, you might find a cart app update that hid the checkout button below a cross-sell carousel on small screens.

Compare channels on how many visitors reach product views and add to cart, not just on purchases. A channel with low purchases but healthy add-to-carts has a different problem than one where visitors leave the landing page immediately. See ecommerce traffic but no sales.

Low product view rates point to navigation, search or category page problems. Check search usage and no-results queries, filter use and exits from category pages. See ecommerce filters.

Compare add-to-cart rates by product and template. Outliers are worth investigating individually. See product page traffic but no sales.

Break the post-cart funnel into its steps to see whether shoppers stall before checkout, at delivery or at payment. See add to cart but no purchase and why customers abandon checkout.

Track repeat purchase rate and time to second order by acquisition channel and first product. A channel that brings many one-time buyers may be worth less than its first-order conversion suggests. See D2C repeat purchase UX.

A drop-off tells you where shoppers leave, not why. Move from symptom to hypothesis to evidence before changing anything. The table below shows common patterns and what to check.

Aggregate funnels hide the problems that matter. Segment by device, traffic source, new vs returning visitors, market and currency, landing page type, product category and, where relevant, logged-in status. A drop that only happens on one device or one market after a release is a bug; a drop across all segments is more likely a pricing or offer issue. See customer segmentation.

Closed funnels only count sessions that start at the first step; open funnels count entries at any step. Checkout often has entries from saved carts or express payment buttons, so compare both. Remember that analytics tools miss some sessions because of consent choices and blockers; compare purchase counts with platform orders and focus on changes rather than absolute gaps. Event definitions matter: see ecommerce event tracking.

The funnel doesn't end at purchase. Delivery problems, returns and poor first experiences reduce repeat purchase. Extend analysis to returns by reason, delivery issues, reviews and repeat purchase by cohort, so conversion improvements that increase returns are visible. See product analytics and customer retention.

The funnel tells you where; it rarely tells you why. Take the leaking stage and segment into qualitative research, such as recordings, heatmaps, surveys and user tests, form hypotheses and test fixes. See ecommerce heatmaps and ecommerce A/B testing. On Shopify, see Shopify conversion funnel optimization and the Shopify funnel audit.

A funnel turns “conversion is down” into “mobile shoppers stop between cart and checkout”. Track every stage, calculate stage-to-stage rates, segment them, and investigate the biggest avoidable drop with qualitative evidence. Then fix, measure and repeat.