UX Health Audit
UX HEALTH AUDIT
August 2026 · UX Health Audit
95 sessions analyzed · 4 journey types · August 11, 2026First report
RELIABILITY
Solid
95 sessions analyzed
SUMMARY
Most users tried to book a seat but never finished — 87 got started, only 24 completed.
What we found
63 sessions tried to book a seat and abandoned — and a handful of people have now tried this multiple times across separate visits without ever completing.
24 sessions completed a booking end-to-end, moving through all five pages in a median of 8s with almost no hesitation at payment.
All 7 sessions checking times and fares completed their goal — this flow works cleanly for browsers.
1 session landed to book and left immediately — not enough to act on yet.
Recommended Actions
Watch and — these are users who got past 'Find trains' and still didn't complete. Look specifically for where they pause on /passenger and what they do before leaving. The passenger details page has the longest dwell time for abandoners (4.7s median) and almost nobody in this group reached 'Continue to payment'.
Query PostHog filtering sessions by the returning user IDs who attempted 'Book a seat' across multiple sessions without completing — a group of repeat abandoners is already visible in the person-level data. Look at which step they exit on each attempt and whether it's consistent. If they're always dropping at the same page, that's your fix.
A dead_click was recorded on 'div#root' at / and a navigation loop was detected on 'northrail.uxsense.ai/' — both in the abandonment group. Watch (the exited session) to see if the homepage interaction failure is reproducible. If div#root is swallowing clicks, some users may never be getting their search submitted.
The booking flow works — completers move through it quickly and without hesitation — but most users are hitting a wall before they get to payment, and a group of repeat visitors have already tried and failed multiple times. Fix what's stopping users on /passenger first.
How Confident
Moderate-to-good — 95 sessions is a reasonable starting point and the abandonment pattern is consistent, but the repeat-struggler picture will sharpen as more sessions come in.
This audit is your baseline. From your next release on, every report compares behavior against these patterns and tells you what your release changed.