10,000 Heatmap Sessions Later: How Shoppers Actually Browse (And Where You're Losing Them)

By Jonathan · Founder, PageGains

Most e-commerce teams optimize based on what they think shoppers do. They add trust badges because they read a case study, restructure their nav because a consultant suggested it, and A/B test button colors because it's easy. Then they're surprised when conversion rates barely move. After analyzing 10,000 heatmap sessions across a range of e-commerce stores — from sub-$1M DTC brands to mid-market retailers — patterns emerge that make a lot of "best practice" advice look like noise. Here's what the data actually shows.

Shoppers Don't Read Your Product Descriptions — They Scan for One Thing

Across nearly every store in the dataset, click maps showed the same behavior on product pages: visitors land, scan the image, glance at the price, and look for a size or variant selector. The body copy — the paragraph you spent two hours writing about materials, craftsmanship, and brand values — gets almost no engagement. Scroll depth on product pages routinely flattened out around 40%, meaning more than half of visitors never even reached the description.

What shoppers are actually hunting for is reassurance that this is the right variant and that it'll arrive in time. That means your most-clicked elements should be your image gallery, your variant selectors, and your delivery estimate — in that order.

What to do: Move your delivery window (e.g., "Order by 3pm — ships today") directly under the Add to Cart button. Cut your product description to 3–5 bullet points focused on fit, material, and use case. Save the brand story for a collapsible section lower on the page. Stop burying the functional information under marketing copy.

The Navigation Menu Gets Clicked Far Less Than You Think

Here's one that surprises almost every merchant: the main navigation — the thing most teams spend weeks debating — accounts for a small fraction of actual browsing behavior on mobile. In mobile sessions, the hamburger menu got tapped by fewer than 15% of visitors in most stores. The majority of browsing happened through on-page links, related product carousels, and the search bar.

Desktop sessions showed slightly more nav engagement, but even there, only 2–3 top-level categories received the majority of clicks. The rest of the nav was effectively invisible.

What to do: Stop optimizing your nav for edge cases. Focus your energy on the paths people actually take — product page to related products, homepage hero to top category, search results to PDP. Make those paths fast and friction-free. If you're running Google Ads to category pages, check whether the on-page browsing experience does more work than your nav ever will. It usually does.

Rage Clicks Cluster Around One Specific Element — Your Size Guide

Rage clicks (repeated taps or clicks on a non-interactive element) showed up consistently in one spot: size guides. Specifically, shoppers tapped the "Size Guide" link expecting a quick overlay, and when it opened a new page or took more than a second to load, they tapped it again. And again.

This matters because size uncertainty is one of the top reasons for cart abandonment in apparel and footwear. If your size guide creates friction instead of resolving it, you're losing customers at exactly the moment they're trying to commit.

What to do: Your size guide needs to be a fast-loading modal or inline accordion — not a new page, not a PDF. Include a "best for" recommendation: "If you're between sizes, go up." Add a comparison to a common reference brand if your category has one. Smaller stores often see a 5–8% lift in add-to-cart rates just from fixing this single element.

The "Add to Cart" Button Gets Ignored When It Looks Like Everything Else

Button blindness is real. In stores where the Add to Cart button used the same visual weight as secondary actions (like "Save to Wishlist" or "Share"), heatmaps showed diffuse clicking — visitors spreading clicks across multiple buttons rather than the primary one. In stores where Add to Cart was visually dominant — higher contrast, slightly larger, clearly the only bold action on the page — click concentration was dramatically tighter and add-to-cart rates were higher.

This isn't a color argument. It's a hierarchy argument. The page needs to make one choice obvious.

What to do: Audit your product page right now. If your Add to Cart button and your secondary actions look similar in size and contrast, that's a conversion problem. The primary action should be unmistakable. Give it a specific label when possible — "Add to Cart — Free Returns" outperforms a plain "Add to Cart" because it handles the risk objection at the moment of decision. One button should dominate the page visually. Everything else should recede.

Mobile Shoppers Behave Completely Differently — And Most PDPs Are Built for Desktop

Mobile sessions made up the majority of traffic in virtually every store analyzed, yet product pages were clearly designed with desktop layouts in mind. On mobile, key information — the price, the variant selector, the CTA — was often pushed below the fold by oversized hero images or auto-playing video. Scroll maps showed that a significant portion of mobile visitors never scrolled past the first image.

Meanwhile, on desktop, the same stores often had too much whitespace and content spread across a wide canvas that made the page feel sparse and unconvincing.

What to do: Design your product page on mobile first, then adapt it to desktop — not the other way around. On mobile, the image, variant selector, price, and Add to Cart button should all be accessible without scrolling. A sticky Add to Cart bar that appears after the visitor scrolls past the fold can recover a meaningful portion of mobile sessions. Test this before anything else if you haven't already.

Visitors Who Use Search Convert at 2–3x the Rate — But Most Search Bars Are Broken

Search users are self-selecting as high-intent. They know what they want. They're not browsing — they're looking for something specific, and if they find it, they're ready to buy. Across the sessions analyzed, visitors who used the search bar had dramatically higher conversion rates than non-search visitors. The problem is that most on-site search experiences are terrible.

Common failures: no autocomplete, no tolerance for misspellings, no handling of synonyms ("couch" vs. "sofa"), and zero-results pages with no fallback recommendations. Shoppers hit a dead end and leave.

What to do: Treat your search bar as a primary revenue channel. Install autocomplete if you don't have it. Set up synonym rules for your top-searched terms. Build a real zero-results page that surfaces best-sellers or suggests alternative searches. Check your internal search analytics monthly — the queries with zero results are a direct map to revenue you're leaving on the table. Fixing search is one of the highest-ROI improvements available to most mid-sized stores.

Social Proof Works — But Only When It's Specific and Placed Right

Every store in the dataset had reviews somewhere. Most placed them at the bottom of the product page. Heatmaps showed that the majority of visitors never scrolled that far. The reviews existed; the visitors never saw them.

The stores that saw review content actually influence behavior had one thing in common: they surfaced specific, concrete proof earlier in the page — a star rating and review count directly under the product title, a pulled quote near the size guide, a UGC photo in the image gallery. Generic "4.8 stars from 2,000 reviews" in the right position outperforms a wall of five-star testimonials buried at the bottom.

What to do: Pull your average star rating and review count immediately under the product name — this should be one of the first things a visitor sees. Pick one specific review that addresses your most common objection (sizing, quality, shipping speed) and surface it near the Add to Cart button. Don't rely on visitors scrolling to find your social proof. Place it where they're already looking.

Exit Intent Is Happening Earlier Than Your Pop-Up Triggers

Most exit-intent pop-ups are configured to fire when the mouse moves toward the browser's close button. On desktop, this catches some users. On mobile, it catches almost none — because mobile exit behavior doesn't translate to that mouse-movement signal. But the bigger issue is that by the time someone is actively trying to leave, the decision is usually already made. Heatmaps showed that disengagement — clicks stopping, scroll reversing, attention drifting — typically happened 20–30 seconds before the actual exit.

Chasing people at the door is less effective than catching them when they first stall.

What to do: Look at where scroll depth plateaus on your key pages. That flat line in your scroll map is where visitors disengage. That's the section you need to fix — a clearer value proposition, a better image, a more compelling proof point. The exit pop-up is a last resort; fixing the content that precedes disengagement is the real opportunity. Use session recordings alongside heatmaps to watch what happens in the 30 seconds before users stop engaging.

The Bottom Line

Most conversion problems aren't about the wrong button color or a missing trust badge. They're about a fundamental mismatch between how shoppers actually move through a page and how the page was built. Shoppers scan, not read. They use search more than navigation. They disengage quietly before they leave loudly.

The heatmap data doesn't lie, but it does require interpretation. A click map showing no engagement on your size guide isn't good news — it might mean visitors can't find it, or gave up trying. A scroll map showing 90% drop-off at the halfway point means half your page is invisible to most visitors. The numbers point you to the problem; your job is to figure out why it's happening.

Fix the highest-traffic, highest-friction points first: the mobile product page experience, the Add to Cart hierarchy, the search bar, and the placement of social proof. These aren't glamorous changes. But they're the ones that show up in revenue, not just in reports.