Weight Loss (BO3-B), GLP-2 and ED funnel pages on myrocky.com audited with heatmap.com thermal click overlays and Microsoft Clarity mobile behavior data for the past 7 days. Recommendations are grounded in what is visibly clicked on each mobile viewport.
Mobile users only. Each step counts unique users who visited. Drop-off is the loss between adjacent steps. Note: /checkout volume includes traffic from ED and GLP-2 funnels.
*Checkout traffic exceeds login because some sessions reach /checkout via ED or GLP funnels. The order-received page recorded zero mobile visits, while one Purchase smart event fired - implying tracking gap or off-mobile completion.
Wrap every window.webkit.messageHandlers[X].postMessage call in a guard. The error fires on /bo3-b, /glp1-offer-hero, /checkout, /wl-pre-consultation, and /wl-pre-consultation-2 every time the page loads inside a Facebook/Instagram/TikTok in-app browser. This single fix unblocks the largest traffic segment on the site.
Bind the height and weight labels to their inputs (htmlFor / id), expand tap area to the full row, and fix the "Cannot read properties of undefined (reading 'value')" handler that fires 19x on /wl-pre-consultation-2 and /wl-pre-consultation. The Continue button likely fails on first tap because of this error.
186 dead on whitespace, 122 dead + 24 rage on "Please check your cart...", 58 dead on Continue16% mobile scroll depth on 969 sessions means roughly 814 users never reach the eligibility CTA. Move the "Am I Eligible?" button, social proof badges, and 1-2 benefit lines into the first viewport. Cut everything else above-fold.
16% avg scroll depth, 969 sessions, 85% drop to next step37 quick-back clicks on the /bo3-b primary CTA in 7 days. Users tap, see the destination, then return immediately. Likely causes: slow next-page LCP on mobile, wrong-state quiz step, or the CTA pointing to /wl-pre-consultation-2 (96% scroll depth confirms confused engagement once they arrive).
37 quick-back clicks on "Am I Eligible?" / 12 on "Learn More" / 9 on "Get Started"/glp1-offer-hero gets 567 mobile users; /glp2-offer-hero gets 0 direct mobile, /glp2-pre-consultation gets 10. Pick one as the paid-ad destination, redirect the other, A/B test from a single page. Today the budget is split across two near-identical pages.
/glp1 567 users vs /glp2 0 direct mobile usersThe "Trusted by ▫,▫▫▫+ Users" trust ticker absorbed 3,080 dead clicks and 2,576 rage clicks (high counts likely include animated/auto-firing rotations - audit the tracking). Either link the ticker to reviews, link benefit copy to expandable detail, or change cursor/visual weight so users do not perceive them as tap targets.
Trust ticker, "Am I Eligible?" affordances, FAQ items all generate dead clicks0 /order-received page views but 1 Purchase smart event fired. Either the order-received URL is different on mobile, or the page loads inside a webview that strips the URL match. Confirm tracking pixel fires on mobile completion before judging mobile checkout performance.
0 /order-received mobile users / 1 Purchase event / 29 Checkout events"Cannot read properties of undefined (reading 'value')" also fires on /glp2-pre-consultation. Same root-cause as WL quiz. One patch likely solves both quizzes.
2 JS errors on /glp2-pre-consultation in 7 days, 5 dead clicks







| Browser | Sessions | Bounce % | Dead clicks | Rage clicks | JS errors |
|---|---|---|---|---|---|
| (blank / WebView / in-app) | 1,627 | 89.3% | 145 | 49 | 93 |
| Chrome (desktop) | 115 | 44.4% | 40 | 0 | 0 |
| MobileSafari | 61 | 63.9% | 6 | 0 | 0 |
| ChromeMobile | 46 | 73.9% | 4 | 0 | 1 |
| Edge | 14 | 50.0% | 6 | 0 | 0 |
| SamsungInternet, Firefox, Safari, Unknown | 12 | 75%+ | 0 | 0 | 0 |
Read this carefully: the blank-browser segment (WebViews from Facebook, Instagram, TikTok, Gmail in-app browsers) is 89% of all sessions and accounts for 100% of rage clicks, 96% of dead clicks, and 100% of JS errors. The native browsers (Chrome, MobileSafari, ChromeMobile, Edge) collectively bounce at 44-74% with virtually zero JS errors. Fix the WebView path and the entire site quality jumps overnight.
Two data sources combined: heatmap.com (real thermal click overlays from the portal canvas, mobile device filter, current month) and Microsoft Clarity MCP (mobile-only, May 1 - May 8 2026, US site myrocky.com). Heatmap canvases are exported from the heatmap.com portal at the page's recorded mobile width and composited onto a fresh Puppeteer mobile screenshot at iPhone 14 dimensions (375 x 812 @ 2x DPR). Click density on the heatmap is the source of truth; Clarity provides counts, scroll depth, JS errors and funnel progression.
Note: low-traffic pages and pages without heatmap.com tracking show the raw mobile screenshot only. Browser breakdown covers all-device sessions to surface in-app browser issues that appear sitewide.