How to Find Conversion Friction Using Funnel + Recording Analysis

The Two-Layer Approach to Finding Friction

Quantitative data (funnel analysis) tells you where users drop off. Qualitative data (session recordings) tells you why. Used together, they produce specific, testable hypotheses rather than vague hunches about "improving UX."

Our conversion optimization team always starts with this two-layer analysis before proposing any test or change.

Layer 1: Funnel Analysis in GA4

GA4's Funnel Exploration report (in the Explore section) lets you define a multi-step conversion path and see the drop-off rate at each step.

Setting Up Your Funnel

1. Go to GA4 → Explore → Funnel Exploration
2. Define your steps using events or page views:
   • Step 1: session_start (or landing page view)
   • Step 2: product_detail_view (or pricing page view)
   • Step 3: add_to_cart / begin_checkout / form_start
   • Step 4: purchase / form_submit
3. Set date range (minimum 30 days for statistical reliability)
4. Apply segment: Users from your primary paid channel (to analyze your most expensive traffic)

Reading the Funnel

The output shows percentage of users completing each step. Your target is the step with the largest absolute drop-off (number of users lost, not just percentage).

Example reading:

• Landing page → Product page: 45% pass through (55% bounce — investigate landing page)
• Product page → Add to cart: 25% pass through (friction at product-to-cart stage)
• Add to cart → Checkout: 65% pass through (normal; some abandonment is expected)
• Checkout → Purchase: 30% pass through (significant friction at checkout)

In this example, both the landing page bounce and checkout abandonment deserve attention. The landing page affects more users (all traffic); the checkout affects buyers who were close to converting (higher value).

Layer 2: Session Recordings for Each Friction Point

For each significant drop-off step, pull session recordings from behavior analytics tools (Clarity, Hotjar) filtered to users who reached that step but didn't proceed.

For "Product page → Add to cart" drop-off

Filter: Sessions that included a product_detail_view event but no add_to_cart event

Watch 30–50 sessions. Look for:

• Do users scroll to the "Add to Cart" button? (If not, it's visibility)
• Do users interact with product images, size selectors, or options? (If yes but no cart, they may be confused)
• Do users click the product title or images expecting more detail? (Dead clicks = missing content)
• Do users visit the shipping/returns page and not come back? (Concern about shipping cost)

For "Checkout → Purchase" drop-off

Filter: Sessions that included begin_checkout but no purchase event

Watch 30–50 sessions. Look for:

• Rage clicks on the payment submit button (broken payment processing)
• Users reaching the payment form and immediately leaving (trust concern — no security indicators)
• Users entering card details, then stopping at billing address (unexpected required field)
• Mobile users struggling with form field layout

Converting Observations to Hypotheses

After analysis, structure findings as: "We observed [X pattern] in [Y% of sessions], which suggests [Z problem], which we can test by [change]."

Example: "We observed 35% of checkout abandoners paused at the billing address field (which requires a UK postcode lookup) for more than 30 seconds before leaving. This suggests the postcode lookup tool is a friction point. We can test by replacing it with a free-text address entry form."

This specific hypothesis — coming from combined funnel + recording analysis — is ready to be A/B tested with a clear expected outcome. Connect test results back to GA4 conversion events to measure impact. Our CRO team runs this analysis systematically. Contact us to identify your highest-impact conversion friction points.