Challenge. Mobile accounted for 72% of traffic but only 41% of bookings; the legacy form took ~4m10s to complete and had a 58% drop-off between quote and payment. Quote requests generated ~120 support emails/week for clarifications (flight numbers, child seats, luggage). Dispatch spent ~18 staff hours/week correcting addresses and job details. Goals: lift mobile conversion, shorten time-to-book, reduce ops overhead, and modernise brand perception.
Solution. We rebuilt the site (Next.js, edge caching) and launched a 3-step Plan → Select → Pay flow. AI/NLP now parses free-text locations to structured addresses with 98.2% match accuracy; an AI recommender boosts relevant vehicle picks by 34%. Dynamic pricing preloads top options; an in-flow AI assistant deflects routine questions. Apple Pay/Google Pay and saved-profile checkout reduced input fields by 43%. Accessibility (WCAG 2.2 AA) and a mobile-first design improved tap targets and readability.
Implementation. Core Web Vitals moved to LCP 1.7s (-48%), CLS 0.03, INP 170ms across the top 20 pages. Real-time distance/time and flight lookups trimmed manual edits; rulesets auto-handle add-ons. A headless CMS enabled 120+ landing pages; schema and internal linking increased non-brand impressions by 61% in 60 days. Webhooks push validated jobs to dispatch/CRM; anomaly detection flags ~3.6% edge cases for review. Synthetic monitoring and dashboards track funnel health and error budgets.
Results (90 days post-launch).
Mobile conversion: +38% (1.84% → 2.54%); desktop +17%.
Time-to-book: -52% (4m10s → 2m00s).
Form errors: -46%; address corrections: -81% (18 → 3.4 staff hrs/week).
Support emails: -63% (120 → 44/week); AI assistant self-serve resolution 74%.
Checkout completion: +29%; Apple/Google Pay adoption 42% of mobile checkouts.
Revenue per session: +22%; refund/adjustment rate -19% from cleaner job data.
Luxcar’s digital experience now matches its premium on-road service, while the AI layer keeps learning to make each booking faster, clearer, and more profitable.