Real-user data through
CrUX rolling 28-day window
Page and vendor scan
Latest usable Lighthouse measurement · over 14 days old
Sitemap checked
Latest recorded sitemap check
Latest observations · Jeff D'Ambrosio Of Oxford Mitsubishi
A closer look at the latest measurements. See what moved, when we observed it, and where the evidence stops.
Real-user performance
1 improved · 2 worsened
3 comparable metrics · overlapping windows may share visits
Vendor observations
5 differences observed
4 newly seen · 1 not seen again
Page checks
1 score changed
Accessibility, SEO, and best practices
These are observed differences, not proof of cause. Vendor changes do not establish why performance moved. Check the dates below: the latest available observations may be old.
CrUX field data · 75th percentile · lower is better
Largest contentful paint
1,323 ms1,250 ms
Interaction to next paint
104 ms106 ms
Cumulative layout shift
0.160.19
Oct 3, 2026 → Oct 10, 2026
Two CrUX rolling windows. Windows may overlap; differences are descriptive, not a significance test.
Lighthouse desktop audits · scores out of 100
Accessibility
89 points89 points
SEO
85 points85 points
Best practices
46 points65 points
Sep 2, 2026 → Sep 30, 2026
Two usable desktop scans. Lab results vary between runs; audit versions are not recorded, so score changes may also reflect tool updates.
Detected in the latest usable scan compared with the previous one. Consent, timing, and missing audit data can affect detections.
Seen in the latest scan, but not in the previous scan.
Tentative: absence in one scan does not confirm removal.
Sep 2, 2026 → Sep 30, 2026
Two usable desktop scans. Lab results vary between runs; audit versions are not recorded, so score changes may also reflect tool updates.
Recorded transfer, requests, and main-thread work. Smaller is generally better; one scan is not a trend.
Page weight
18.7 MB18.43 MB
Page requests
315 requests318 requests
Main-thread work
16,027 ms9,004 ms
Sep 2, 2026 → Sep 30, 2026
Two usable desktop scans. Lab results vary between runs; audit versions are not recorded, so score changes may also reflect tool updates.
