Carepatron
Onboarding redesign · 22 May 2026

Four bets on what should replace the Get Started dashboard.

The current 6-step checklist is failing. 53.5% of new Americas signups never click it, and the users who do activate (24% create a client) mostly do so by bypassing the checklist entirely. Each prototype below tests a different hypothesis about what should sit at post-signup instead. Open each, walk through the flow (around 30 to 60 seconds), then read the side-by-side at the bottom.

Never engage
53.5%
never click the dashboard
Create a client
24.0%
real activation, often skips dashboard
Connect Stripe
5.1%
monetisation gate
Subscribe
4.4%
Americas cohort, Apr 15 to May 7
Prototype 01

Action-first

The bet. Strip away everything else. The product is the onboarding. Sign up, create a client, run an appointment, get paid.

No dashboard. Post-signup lands on "Create your first client." Each action unlocks the next contextual nudge: client, then appointment, then Stripe. Setup tasks (location, branding, team) defer to a slide-in "Practice setup" drawer that tracks against the old 6-task list, so the legacy pattern stays preserved without competing.

7 steps Around 45 seconds Lowest eng cost
Open prototype
Prototype 02

Switcher vs greenfield

The bet. Segmentation is the unlock. Switchers (SimplePractice migrators) and greenfield therapists have radically different first jobs. Serve them separately.

One question at the door: "Switching, or starting fresh?" Switcher path leads with CSV import and tool-specific migration tips for SimplePractice, TherapyNotes, and Jane. Greenfield path leads with create-first-client. Both converge on a shared "you're ready" state, but the velocity differs. Scale-first vs depth-first.

2 paths Around 60 seconds Best fit for switchers
Open prototype
Prototype 03

Ava-conversational

The bet. AI conversation scales to context better than a static checklist, and no competitor (Jane, SimplePractice, TherapyNotes) does this. A real differentiator.

Ava is the onboarding. Full-screen chat where Ava asks questions and renders inline forms inside the conversation. Client name leads to appointment leads to Stripe through dialogue. The dashboard only reveals itself after the chat collapses to a corner bubble. Completion feels like Ava built the practice for you.

Around 30 seconds Differentiator Highest eng cost
Open prototype
Prototype 04

Value preview

The bet. Tangibility unlocks intent. Show a working practice with 3 fake clients before asking the user to build one. "Make it yours" claims the workspace.

Lands on a fully populated demo: 3 clients, a real calendar, draft and paid invoices, two SOAP notes. All 5 nav tabs are lived-in. "Make it yours" in the nav opens a claim flow. Demo data fades, the user's first real client takes its place, Stripe toast follows. Show, don't tell.

5 tabs explorable Highest wow Hardest to build
Open prototype

Side by side

How each prototype handles the same problem.

01 Action-first 02 Switcher / greenfield 03 Ava-conversational 04 Value preview
Mental model "Just do the thing" "What's your job?" "Let me guide you" "Here's the value"
Activation goal First client created First client or import done First client and first appt "Claim" the workspace
Best for Greenfield therapists Switchers especially Low-confidence new users Tire-kickers and evaluators
Weakness Switchers feel ignored Misroute equals wrong path Power users may chafe "Demo data" objection
Reversible Yes, A/B test Yes, A/B test Partial, Intercom config Harder, new product surface
Eng cost Low, around 2 weeks Medium, around 3 to 4 weeks Medium, around 3 weeks High, around 6 to 8 weeks

If I were placing bets

Cheapest tests first, biggest swings last.

  1. Ship 02 (switcher vs greenfield) as a 50/50 A/B test against the current dashboard. Lowest risk, highest information value. Confirms whether persona-split alone moves create-client rate. Primary metric: create-client-in-7d. Around 2-week build, 3-week runtime.
  2. Test 01 (action-first) as a second arm. Once #1 is running, fork the greenfield path into action-first. Tests whether the dashboard is needed at all for that persona. A cheap follow-on.
  3. Stand up 03 (Ava) as a parallel track. Most of this is Intercom config, and the Ava audit playbook is already documented in fin-performance-audit. Doesn't need eng resources. Can run alongside the A/B test.
  4. Treat 04 (value preview) as a Q3 bet, conditional on what 01 and 02 reveal. It's the highest-ceiling concept but also the most expensive. Earn the right to build it by killing the easy options first.