At a glance
LCP side-by-side mobile vs desktop, surfacing the device-specific load-speed gap. The LCP gap is the largest contributor to the overall mobile-vs-desktop score gap because LCP is heavily network-dependent and mobile networks are slower. Typical healthy gap: 1,000-2,500ms (desktop faster). Larger gaps signal mobile-specific image / network issues; smaller gaps suggest a well-optimised mobile experience.
Calculation
Calculated automatically from your Website Performance (PageSpeed + CrUX) data. See the At a glance summary above for what the metric tracks and the worked example below for a typical reading.Worked example
A UK-based BigCommerce fashion store, lab LCP per template Wednesday 15 May 26.
What the gap analysis is telling us:
-
Homepage gap of 2,900ms is the dominant problem. The hero PNG carousel (4 images at 2.8MB total) downloads quickly on desktop wifi but takes 12+ seconds on slow 4G mobile. The fix: responsive variants via
srcsetso mobile downloads 400KB-sized variants instead of 2.8MB full-resolution. Estimated post-fix mobile LCP: 2,400-2,800ms; gap closes to 800-1,000ms. - PDP and collection gaps in 2,100-2,400ms range are typical-to-borderline. Same fix-pattern: image format + responsive variants. Combined fix lifts mobile LCP across all three high-traffic templates.
- Cart and checkout have healthy 760ms gaps. No mobile-specific work needed; transactional pages are already lean.
- The aggregate gap of 2,200ms is right at the borderline between “typical” and “concerning”. The high-traffic templates (homepage, PDP, collection) drive the gap; the lighter pages (cart, checkout) don’t contribute meaningfully.
- Recovery target: close the gap to 800-1,200ms across all templates. Combined fix-cycle: image responsive variants on hero + product + collection images. Estimated 2-3 weeks of focused work; lifts mobile LCP from 3,920ms toward 2,200-2,400ms (good band).
- Identify templates with concerning gaps (>2,500ms or >3,000ms). Concentrate there.
- Decompose by LCP element type. Hero image gap = image fix; hero text gap (rare) = render-blocking fix.
- Apply mobile-specific image work: responsive variants via
srcset,loading="eager"on LCP image,<link rel="preload">.
Sibling cards merchants should reference together
Reconciling against the vendor’s own dashboard
Where to look:- PageSpeed Insights, toggle Mobile / Desktop tabs; LCP appears in field section.
- Chrome DevTools → Lighthouse panel, choose device profile.
Cross-connector reconciliation: primarily internal (with
psi_mobile_vs_desktop_score, crux_lcp_p75).
Quick rule for support tickets: if a merchant says “my LCP gap is 4 seconds, what now?”, the answer is responsive image variants via srcset. The gap is almost always image-driven on ecommerce sites.
Known limitations / merchant FAQs
Why is my mobile LCP 4 seconds slower than desktop? Mobile-specific image issue, almost always. Hero images served at desktop resolution (1920px wide) take 14+ seconds to download on slow 4G. Addsrcset responsive variants so mobile downloads 400-800px-wide versions instead.
Will responsive variants fix this?
Yes, dramatically. Typical mobile LCP improvement from srcset implementation: 1,500-2,500ms. The gap to desktop closes from 3,000ms+ to 800-1,200ms.
My BC Stencil theme says it handles responsive images. Is that enough?
Verify in DevTools Network. Theme defaults sometimes don’t include enough breakpoint variants for modern device sizes. Check: are there variants at 400px, 600px, 800px, 1200px widths? If only 1920px exists, mobile gets the desktop image.
Should I optimise mobile LCP to match desktop LCP?
Approximately. Target mobile LCP under 2,500ms (Google’s good threshold). The gap to desktop’s typical 1,500ms is structural and acceptable. Closing past 1,000ms gap requires architectural-level work usually not justified.
Can image-CDN handle this automatically?
Yes, if configured. Cloudinary, imgix, BC’s image CDN handle format negotiation + responsive resizing automatically when the theme HTML uses appropriate srcset. The merchant’s responsibility: ensure the theme HTML opts in.