At a glance
Per-template Mobile vs Desktop score comparison, surfaces the device gap broken out by page type (homepage, PDP, collection, cart, checkout, etc.). Different templates carry different mobile-vs-desktop gap patterns; this card lets the merchant identify which templates need targeted mobile work. Pairs with psi_mobile_vs_desktop_score which gives the site-weighted aggregate; this card decomposes by template.
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, per-template comparison Wednesday 15 May 26.
What the per-template view is telling us:
- Collection page has the highest gap × traffic score (8.7). Despite homepage having the largest gap, collection has higher traffic share, making it the highest-leverage optimisation target. Fix collection page first.
-
The optimisation priority by gap × traffic:
- Collection page (8.7): filter widget refactor + responsive images on grid items
- Homepage (7.5): hero image responsive variants
- PDP (6.0): hero image responsive variants
- Cart, search results (2.1 each): low priority
- Checkout, about (sub-1.0): no work needed
-
Same fix-pattern across templates for image-heavy pages: responsive image variants via
srcset, image format conversion, preload critical images. One workflow addresses homepage + PDP + collection-page-grid simultaneously. -
Collection page additional work: filter widget refactor (covered in
psi_mobile_desktop_inpplaybook). The mobile-specific INP issue compounds with the LCP issue on collection. -
Realistic post-optimisation forecast for the worst-3 templates:
- Collection: mobile 48 → 70 (gap from 31 → 9)
- Homepage: mobile 42 → 65 (gap from 34 → 11)
- PDP: mobile 58 → 75 (gap from 24 → 7)
- Site-weighted mobile score: 51 → 68
-
Reading the trio with
psi_mobile_desktop_lcp,psi_mobile_desktop_inp,psi_mobile_desktop_clsdecomposes which sub-metric drives each template’s gap. Collection’s gap is INP-dominated; homepage’s is LCP-dominated. Different fix focus per template.
- Sort templates by gap × traffic.
- Decompose dominant template gaps by sub-metric using the per-CWV device-gap cards.
- Apply template-level fixes: each template’s fix benefits all URLs sharing the template.
- Re-audit to confirm gap reduction.
Sibling cards merchants should reference together
Reconciling against the vendor’s own dashboard
Where to look:- PageSpeed Insights, paste each template’s representative URL; toggle Mobile/Desktop tabs.
- Lighthouse CI, runs per-URL audits in build pipeline.
Cross-connector reconciliation: primarily internal (with the per-CWV device-gap cards and per-URL ranking cards).
Quick rule for support tickets: the gap × traffic ranking is the actionable view. Don’t optimise the template with the largest gap if its traffic share is small; optimise the template with the largest gap × traffic product.
Known limitations / merchant FAQs
Should I optimise the template with the largest gap or the largest gap × traffic? Largest gap × traffic. The product determines commercial impact: a 35-point gap on a low-traffic page affects fewer users than a 24-point gap on a high-traffic page. Why does collection page have a larger gap than homepage? Filter widget JS. Collection pages run mobile-specific filter logic that desktop doesn’t trigger (desktop usually has always-visible filter sidebar). The mobile-only widget compounds the structural CPU gap. Will fixing one template improve other templates? Some patterns transfer. Image optimisation in the homepage hero often re-uses the same image-CDN config that benefits PDP and collection images. Filter widget refactor benefits all collection pages simultaneously. Template fixes have multiplier effects. Can Vortex IQ tell me which sub-metric drives each template’s gap? Yes; cross-reference withpsi_mobile_desktop_lcp, psi_mobile_desktop_inp, psi_mobile_desktop_cls. Each surfaces the per-template breakdown for its sub-metric.