Skip to main content

Starter Voice — Founder Interview Answers

Status: COMPLETE — All 20 questions answered. Date: 2026-03-28 Interviewer: Claude (Opus 4.6)

Section A: Onboarding & First Email​

Q1: Welcome Email Tone (No Trial)​

Decision: Two-email provisioning flow. Voice is NOT instant like chat — there's a real provisioning step (Telnyx number, SIP trunk, LiveKit dispatch rule) with internal QA.

Email 1 (immediate post-checkout):

  • Tone: "Welcome! We're getting your dedicated church phone line ready."
  • Step 1: "That's on us — we're provisioning your phone line and running quality checks on your voice agent. You'll hear from us within 1 business day with your number."
  • Steps 2-4: "While you wait, get things ready on your end:" — church profile, denomination, voice agent personality, notification preferences
  • Includes dashboard magic link so they can start setup immediately

Email 2 (after QA, within 1 business day):

  • Personalized — checks what customer has completed during wait
  • Celebrates progress: "Great job setting up your service times and staff!"
  • Nudges remaining items: "Here are a few more things that will make your voice agent soar!"
  • Includes actual phone number
  • Includes carrier-specific call forwarding instructions (based on phone provider selected during onboarding)
  • "Make a test call!" CTA

Q2: Welcome Email "Things to Do First"​

Decision: Steps 1-4 + optional step 5, mirrored in the dashboard Getting Started checklist:

  1. (On us) Getting your phone line and voice agent ready
  2. Complete your church profile (service times, staff, ministries)
  3. Set your denomination/theological tradition
  4. Customize your voice agent's personality
  5. Set up your notification preferences

Q3: Annual Billing Toggle​

Decision: No annual billing for voice plans. Toggle does not appear when voice channel is selected on pricing page, or anywhere in the voice dashboard.

Additional decision: Add product_knowledge FAQ explaining why no trial/annual for voice — complexity and cost of voice lines, keeping prices low for all churches.

Section B: Dashboard — Tab Visibility​

Q4: Starter Voice Tabs​

Decision: 6 tabs:

  1. Overview
  2. Calls
  3. Requests
  4. Training
  5. Settings
  6. Upgrade

Hidden:

  • Care (not applicable for voice-only — backlogged for future outbound calling feature)
  • Website (no PewSearch bonus at Starter)
  • Social (separate product)

Principle: Chatbot is completely invisible in the voice-only dashboard — same as voice is invisible in chat-only. Hidden, not locked.

Q5: Chatbot References​

Decision: Zero reference to chatbot anywhere in the voice-only dashboard. Completely invisible, same principle as hiding voice from chat-only customers.

Section C: Overview Tab​

Q6: Overview Stat Cards​

Decision: 5 stat cards:

  1. Total calls this month
  2. Prayer requests captured
  3. Visitor contacts captured
  4. Callback requests
  5. Average call duration

No "Active care members" (Care tab is hidden for voice-only).

Q7: Getting Started Checklist​

Decision: 5 steps, mirrors the welcome email. Auto-dismisses after 7 days or all complete (same as Starter Chat).

  1. Complete your church profile (hours, staff, ministries)
  2. Set your denomination/theological tradition
  3. Customize your voice agent's personality
  4. Forward your church phone number (only actionable after "line is live" email)
  5. Make a test call

Q8: Upgrade CTA on Overview​

Decision: Same "Did you know?" subtle educational moments as Starter Chat. No buttons, just text linking to Upgrade tab. Voice-flavored: e.g., "Did you know you could add a chatbot to go along with your voice agent? They make quite the team!"

Section D: Training Tab​

Q9: Training Sub-Tabs​

Decision: Same 5 sub-tabs as Starter Chat:

  1. Church Knowledge
  2. This Week
  3. Theology
  4. Agents (with voice-specific controls: voice picker, greeting config)
  5. Safety

Hidden: FAQs (Pro+), Simulator (Pro+)

Voice agent uses the same church knowledge (hours, staff, ministries, denomination, this week's sermon) as the chatbot. Training content is shared.

Q10: Agents Sub-Tab (Voice-Specific)​

Decision: Same 2 agents visible (Care + Coordinator; Discipleship/Stewardship hidden). Personality editor same fields plus voice-specific controls (voice picker from Cartesia, greeting script configuration).

Q11: Testing Voice Agent​

Decision: No in-dashboard simulator for voice. Pastor calls the actual number to test. Call logs support soft delete/hide so test calls can be cleaned from the dashboard (record flagged in DB, not actually deleted — preserves legal accountability).

Section E: Settings Tab​

Q12: Settings Sections​

Decision: Same as Starter Chat (Church Profile, Hours, Notifications, Integrations, Team) with these voice-specific additions:

  • Phone number display in Overview tab header or Church Profile section (read-only, shows assigned AI number)
  • No embed widget settings (no chatbot)
  • No chatbot widget color/branding settings
  • No call forwarding instructions in dashboard — those are delivered via the "Your line is live!" email with carrier-specific steps

Q13: "Powered by ChurchWiseAI"​

Decision: Same pattern as chat. "Powered by ChurchWiseAI" branding visible in the admin dashboard (admin-facing only, not caller-facing). Callers hear AI disclosure in the voice greeting — that's the caller-facing equivalent.

  • Starter/Pro: Cannot remove branding
  • Suite: Can remove (white-label)

Q14: Team Member Limits​

Decision: Same as Starter Chat — unlimited team members, 9 roles. No reason to differ.

Section F: Upgrade Tab​

Q15: Upgrade Tab Options​

Decision: Mirror of Starter Chat upgrade tab but voice-flavored. Shows comparison cards for Starter/Pro/Suite voice tiers plus an "Add Chat" section. Feature comparison table, same format.

Section G: Calls Tab​

Q16: Call Log Columns​

Decision: Expandable row pattern (same as Requests tab):

Table row (scannable):

  • Date/time
  • Duration
  • Caller phone number (or "Unknown")
  • Intent detected (prayer request, general inquiry, callback, etc.)
  • Status badge (new / reviewed)

Click to expand:

  • Full transcript (no audio recording — transcript only for now)
  • Which agent handled it (Coordinator, Care, etc.)
  • Actions taken (prayer request captured, callback scheduled, etc.)
  • Soft delete/hide button

Q17: Call Recordings​

Decision: No audio recordings for now. Transcript only. Keeps things simpler, cheaper, and avoids recording consent issues.

Section H: Requests Tab​

Q18: Requests Tab UI​

Decision: Same UI as chat Requests tab. Same 3 types: prayer requests, callback requests, visitor contacts. Data sourced from voice calls instead of chat. Shows caller phone number where chat shows "website visitor."

Section I: Lifecycle & Billing​

Q19-Q20: Cancellation & Failed Payment​

Decision: DEFERRED to the Cancelled and Trial Expired interviews. Those specs will cover all tiers (including voice-specific scenarios like: what happens to callers when subscription ends, grace periods, what callers hear). Noted in interview-action-items.md.

Additional Notes​

  • Onboarding form needs "Who is your phone provider?" field — used to generate carrier-specific forwarding instructions in Email 2
  • Soft delete applies to ALL record types (call logs, prayer requests, visitor contacts, callbacks) across ALL tiers, not just voice
  • Care tab for voice is backlogged — will appear when outbound calling feature is built (voice agents calling opted-in congregants)