> ## Documentation Index
> Fetch the complete documentation index at: https://docs.vortexiq.ai/llms.txt
> Use this file to discover all available pages before exploring further.
# Load Speed: Mobile vs Desktop, Website Performance (PageSpeed + CrUX)
> Load Speed: Mobile vs Desktop for Website Performance (PageSpeed + CrUX) stores. Tracked live in Vortex IQ Nerve Centre. How to read it, why it matters, an...
**Metrics type:** [Supporting Metrics](/nerve-centre/overview#metrics-types-explained) • **Category:** [Website Performance](/nerve-centre/connectors#connectors-by-type)
## 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.
| | |
| ----------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **What it counts** | Two LCP measurements from the same URL, one per device profile. Reported side-by-side with the gap. Each measurement is a single Lighthouse audit run per device. |
| **Sample type** | **Lab data** from two Lighthouse runs. Field equivalents available via [`crux_lcp_p75`](/nerve-centre/kpi-cards/website-performance/page-load-speed) (mobile + desktop separately). |
| **Why mobile LCP is structurally slower** | (1) **Slow 4G throttling** (1.6 Mbps down vs unthrottled desktop), for a 2.8MB hero image, mobile takes \~14 seconds to download vs \~0.5 seconds on desktop. (2) **Different LCP element selection**: smaller mobile viewport may select a different "largest" element (often the hero image where desktop selects the header text). (3) **Less browser cache benefit**: mobile users more often hit cold cache. |
| **Healthy gap interpretation** | **0-800ms gap**: site is well-tuned for mobile; structural emulation differences explain the gap. **800-2,500ms gap**: typical for ecommerce. **2,500-4,000ms gap**: meaningful mobile-specific issues (oversized images served to mobile). **4,000+ms gap**: mobile-broken; mobile users experience seriously degraded load times. |
| **Reading the gap diagnostically** | Large LCP gap usually traces to image-related issues: hero images served at desktop resolution to mobile viewports, missing `srcset` responsive variants, no `loading="eager"` + `` for the LCP element. **Fix is mostly mechanical**: image format conversion + responsive variants closes the gap dramatically. |
| **Currency** | n/a, two milliseconds + delta. |
| **Time window** | `T/7D` |
| **Alert trigger** | `gap > 3,000ms` (mobile-broken signal). |
| **Sentiment key** | `psi_lcp` |
| **Roles** | owner, marketing, operations |
## 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.
| Page template | Mobile LCP | Desktop LCP | Gap | Read |
| ------------------------- | ----------: | ----------: | ----------: | ------------------------------------------------ |
| Homepage | 4,820ms | 1,920ms | **2,900ms** | Concerning; hero image not served at mobile size |
| Product detail page | 3,640ms | 1,540ms | 2,100ms | Typical |
| Collection page | 4,210ms | 1,840ms | **2,370ms** | Borderline |
| Cart | 2,180ms | 1,420ms | 760ms | Healthy |
| Checkout | 1,940ms | 1,180ms | 760ms | Healthy |
| **Site-weighted average** | **3,920ms** | **1,720ms** | **2,200ms** | **Typical-to-borderline** |
What the gap analysis is telling us:
1. **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 `srcset` so 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.
2. **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.
3. **Cart and checkout have healthy 760ms gaps.** No mobile-specific work needed; transactional pages are already lean.
4. **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.
5. **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).
The diagnostic flow:
1. **Identify templates with concerning gaps** (>2,500ms or >3,000ms). Concentrate there.
2. **Decompose by LCP element type.** Hero image gap = image fix; hero text gap (rare) = render-blocking fix.
3. **Apply mobile-specific image work**: responsive variants via `srcset`, `loading="eager"` on LCP image, ``.
Rapid-response playbook:
| Time horizon | Action |
| -------------- | ---------------------------------------------------------------------------------------------------------------- |
| First 1 hour | Identify worst-gap templates. |
| First 24 hours | Add `srcset` responsive variants to hero images. |
| First 7 days | Mobile LCP improvement visible in lab. |
| Day 28 | Field LCP via [`crux_lcp_p75`](/nerve-centre/kpi-cards/website-performance/page-load-speed) reflects the change. |
## Sibling cards merchants should reference together
| Card | Why merchants reach for it |
| ---------------------------------------------------------------------------------------------------------- | ------------------------------------------------------ |
| [`psi_mobile_vs_desktop_score`](/nerve-centre/kpi-cards/website-performance/mobile-vs-desktop-score) | Composite score device gap. |
| [`psi_mobile_desktop_inp`](/nerve-centre/kpi-cards/website-performance/responsiveness-mobile-vs-desktop) | INP device gap. |
| [`psi_mobile_desktop_cls`](/nerve-centre/kpi-cards/website-performance/layout-stability-mobile-vs-desktop) | CLS device gap. |
| [`psi_mobile_desktop_ttfb`](/nerve-centre/kpi-cards/website-performance/server-speed-mobile-vs-desktop) | TTFB device gap. |
| [`crux_lcp_p75`](/nerve-centre/kpi-cards/website-performance/page-load-speed) | LCP field measurement (mobile by default). |
| [`psi_image_optimisation`](/nerve-centre/kpi-cards/website-performance/image-optimisation) | Image responsive-variants is the dominant LCP gap fix. |
| [`psi_lab_lcp`](/nerve-centre/kpi-cards/website-performance/lab-load-speed) | Lab LCP measurement. |
## 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.
**Why the Vortex IQ device gap may differ from PSI:**
| Reason | Direction | What to do |
| ---------------------------------------------------------------------------------------------------------------- | ---------------- | ---------------------------------- |
| **Run-to-run variance.** Each profile has independent variance ±10-20%. | Either direction | Use 7-day rolling. |
| **Throttling profile differences.** Lighthouse uses canonical profiles; some local audits use custom throttling. | Small | Confirm canonical profiles in use. |
**Cross-connector reconciliation:** primarily internal (with [`psi_mobile_vs_desktop_score`](/nerve-centre/kpi-cards/website-performance/mobile-vs-desktop-score), [`crux_lcp_p75`](/nerve-centre/kpi-cards/website-performance/page-load-speed)).
**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. Add `srcset` 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.
***
### Tracked live in Vortex IQ Nerve Centre
*Load Speed: Mobile vs Desktop* is one of hundreds of KPI pulses Vortex IQ tracks across Website Performance (PageSpeed + CrUX) and 70+ other ecommerce connectors. Nerve Centre runs the detection layer; Vortex Mind investigates the cause when something moves; Ask Viq lets you interrogate any number in plain English.
[Start for free](https://app.vortexiq.ai/login) or [book a demo](https://www.vortexiq.ai/contact-us) to see this metric running on your own data.