Funnel Data Is Half the Story
Your GA4 funnel shows drop-off points but no causes. You know 32% of visitors add to cart but only 8% complete checkout. That's a 75% cart abandonment rate. But why are they leaving?
Session replay answers the "why." You watch those 32% who made it to cart. You see them scroll past shipping cost and bounce. Or they fill in their address and leave before payment. Or they see the final total and close the tab.
The workflow is simple. Find the leakiest moment. Watch replays of people who hit that moment. Document what you see. Build a fix list. The 80/20 of session replay work lives here.
Step 1: Identify Your Leakiest Micro-Conversion
Pull your funnel report from GA4. Look for the biggest percentage drop between adjacent steps. Not overall drop. Step-to-step drop.
Homepage to collection view. 60% of visitors go there. Collection to PDP. 45% of those continue. PDP to add-to-cart. 28% add to cart. Add-to-cart to checkout start. 12% start checkout.
Your leakiest micro is PDP to cart. You're losing 73% of your PDP visitors. Watch that step first. Fix it first. It compounds.
Step 2: Filter Replays to That Moment
Set up your replay tool to filter for the specific moment. If you're studying PDP-to-cart drop, filter for: visitors who viewed a product page but did not add to cart within 5 minutes. Or filter the opposite: visitors who did add to cart, watch those first to see what converted ones did differently.
Microsoft Clarity makes this easy. You tag events. Then you filter replays by event. Segment it further by device. You're looking at phone replays? Desktop? Both show different friction.
The magic window is 10-20 replays. After 10 sessions, patterns emerge. By 20, you've confirmed them. More than 20 and you're in the weeds with noise.
Step 3: Watch for These Patterns
Take notes on every session. What are you seeing over and over?
Rage clicks: Visitor clicking the product image expecting a lightbox. Or clicking text thinking it's expandable. Document what they clicked.
Dead clicks: Button clicks that register but don't fire. Checkout button not working on mobile. Document the element and device.
Scroll past key content: Visitor scrolls past trust signals, reviews, or shipping info. They never saw it. Change the layout so it's above the fold.
Form field abandonment: They fill in three fields then bounce before completing the form. Which field did they leave at? Maybe it asks for info they don't have. Or the field is confusing.
Heatmaps add the layer. You see where visitors are clicking, scrolling, and spending time. A heatmap shows that 78% of visitors scroll to 40% of the page and stop. Combined with session replay, you know why they stopped.
Step 4: Tag Patterns and Count Frequency
After 20 replays, your notes will show themes. Maybe 8 visitors had the same friction. Maybe 3 visitors did.
Tally by pattern. "Product image not clickable: 7 sessions." "Shipping cost surprises: 5 sessions." "Form field labeling confusing: 3 sessions."
Your fix list is now priority-ordered. Fix the 7-session pattern first. Then the 5. Then the 3.
Want this mapped for your store?
I'll watch 20 replays on your drop-off moment and give you the fix list.
Step 5: Know When to Stop Watching
Stop when the pattern is clear. If 8 out of 15 sessions show the same friction, you've got signal. You don't need to watch all 20.
But if the sessions are all different, watch closer to 20. You're looking for the forest, not individual trees.
The workflow is: watch, document, pattern-match, stop, build fixes. Rinse and repeat for your next leakiest moment.
The Workflow Compounds
Most stores run this once and then stop. They fix the PDP-to-cart issue. Conversion improves. They move on.
The real operators run this monthly. They fix the 1st leakiest moment. Then identify the 2nd leakiest. Watch 20 replays. Build fixes. Ship them. Measure impact. Move to the 3rd.
Three fixes a month, compounding. That's how a 2% conversion rate becomes 4% in six months. Not one big change. Dozens of small changes driven by what actual visitors do.