Message match
The live hero says “Find a psychiatrist,” but the real step is a ten-minute assessment that checks coverage and matches you. I aligned the headline and CTA to the actual job: check coverage, then get matched.
Concept by Kevin Spence for the Senior Manager, Web Experience case study. Not affiliated with Talkiatry. Public stats are illustrative. This shows where my priority experiments point; in practice they ship incrementally as measured tests, not a single redesign. See the rationale →
In-network psychiatry, 100% online
Most patients pay around $30 a visit. Check your coverage in under a minute, then get matched with the right psychiatrist for you.
*Based on patient responses to PHQ-8 or GAD-7 health questionnaires from 1/24–9/25. †Out-of-pocket cost varies by plan and coverage.
“The doctor was extremely empathetic and listened and understood me.”
Teah F., Ohio · Talkiatry patient review
Three quick questions. You will know if we are a fit before you spend ten minutes on the full assessment. No account needed.
Confirm we take your plan and that we treat what you are dealing with, up front.
A short assessment matches you with psychiatrists who fit your needs and preferences.
Book a video visit, often within a week. No waitlists, no commute.
We would rather tell you in one minute than after ten. Here is who we can help today.
A note for the hiring team. This prototype is one focused bet from my 30-day plan: make eligibility clear up front so more qualified visitors start the application, and fewer unqualified visitors stall inside a ten-minute assessment.
The live hero says “Find a psychiatrist,” but the real step is a ten-minute assessment that checks coverage and matches you. I aligned the headline and CTA to the actual job: check coverage, then get matched.
“Qualified application start” is the metric. The eligibility check filters out who we cannot serve (Medicaid, under 5, in-person conditions) while reassuring who we can. That lifts the quality of starts and downstream booking and show rates, not just raw clicks.
A CTA is a promise about what's on the other side of the click, and conversion is a chain of those promises kept. Contextual CTAs are fine; accuracy is the point. Several live buttons ("Book a visit," "Find a psychiatrist") promise a different action than the real next step, a 10-minute assessment, and a broken promise introduces distrust. My primary CTA describes what actually happens, and I'd A/B it against the control. Proof sits next to the CTA, not buried lower.
Public proof (reviews, the 86% feel-better stat, ~$30 out of pocket, in-network with major insurers, compliance badges) is placed at the decision point, with substantiation footnotes. A 988 crisis line is always visible, which a responsible mental-health page should do.
Semantic landmarks, a skip link, labeled form controls with fieldset and legend, visible focus states, AA contrast, large tap targets, and reduced-motion support. Toggle keyboard navigation and it all works.
Open the Events panel (bottom left). Every interaction fires a named GA4-style event with parameters, including the cross-domain assessment_start that is usually where attribution breaks. This is the tracking plan, demonstrated.
Full write-up: audit, prioritized backlog, experiment brief, tracking plan, and QA checklist available on request. Built as a single dependency-free page so it loads fast and every line is reviewable.