Berlein Psychotherapy — metrics
Source of truth for every number on /work/berlein, the homepage berlein chip, and
the berlein row of /repaint. Policy: nothing invented — a row ships to the site
only once it has a real value + source here (danielslater-dev/HANDOVER-HOOK-BEAT.md:114,
UI-BUILD-BRIEF.md:250).
- Site: berleinpsych.com (Vercel).
- Analytics installed: NONE (checked 2026-07-19) — so there is no historical bounce/enquiry/conversion data to recover. Two honest sources only: (a) what we can measure about the site itself now, (b) Alexa's own manual records.
Updated 2026-07-19.
| Metric (site row) | Before | After | Source | Status |
|---|---|---|---|---|
| New enquiries / month | — | — | Alexa (inbox / booking calendar) | ask Alexa (DATA-REQUEST.md); "before" may not exist |
| no credible before (old site; Wayback unreliable) | 98/100 · LCP 2.1s (live rebuild) | Lighthouse mobile, 2026-07-19 | DROPPED — not a before→after (strong after, no measured before) | |
| Mobile bounce | — | — | analytics | no historical data — install CF Web Analytics now, accrues forward |
| No-shows / month | — | — | Alexa (calendar) | ask Alexa; reminders are the mechanism |
| After-hours enquiries captured | — | — | Alexa | ask Alexa |
Honest read: most before→after rows have no "before" — nothing was tracked on the old site. Until real data exists, the options for the case study are: keep it qualitative, show after-only, or omit the table. Load time was dropped (2026-07-19): the rebuild is genuinely fast (98/100, LCP 2.1s) but there's no measured "before" to compare it to, so it's an after-only fact, not a before→after. Do not fill the rest with estimates.