Onboarding Persona Scaling

This drill takes a working persona-based onboarding system (2-3 personas, proven at Baseline) and scales it to handle 5+ personas across multiple channels without proportional manual effort. The key shift: automated persona classification replaces manual tagging, and each persona's onboarding path spans in-app tours, email sequences, and contextual messages simultaneously.

Prerequisites

• onboarding-personalization drill completed with 2-3 working persona tours
• Baseline activation data showing per-persona rates for at least 4 weeks
• At least 500 total users onboarded to have enough data per persona
• Intercom, PostHog, Loops, and n8n all operational

Steps

1. Analyze activation data to discover new personas

Using the posthog-cohorts fundamental, cluster users who activated successfully but do not fit your existing 2-3 persona definitions. Look for patterns in:

• Job title / role (from signup or enrichment)
• First actions taken in the product (sequence of events in first 24 hours)
• Features used most in first week
• Company size and industry
• Signup source and referral path

Identify 2-3 additional persona segments where the current generic fallback tour underperforms. Each new persona must have at least 50 historical users and a measurably different activation path from existing personas.

2. Build automated persona classification

Using n8n-triggers and n8n-workflow-basics, create a classification workflow that fires on signup_completed:

1. Receive the PostHog webhook with signup data
2. Collect signals: role field from signup form, email domain (enriched via Clay if available), company size, signup source UTM parameters, referral path
3. Apply classification rules (if/else chain or weighted scoring):
   • Role contains "engineer" or "developer" AND signup source is "github" or "docs" → technical_builder
   • Role contains "manager" or "director" or "VP" AND company_size > 10 → team_lead
   • Company_size = 1 or role contains "freelance" or "solo" → solo_creator
   • Role contains "marketing" or "growth" → marketer
   • Role contains "sales" or "revenue" → sales_user
   • Default → general (gets the best-performing generic tour)
4. Write persona_type to both PostHog (via posthog-custom-events) and Intercom (via intercom-user-properties API call in n8n)
5. Write persona_confidence property: "explicit" (user selected role), "inferred" (classification rules), or "default" (no signal)

3. Create per-persona email sequences

Using the loops-sequences fundamental, build persona-specific email sequences alongside the in-app tours. Each persona gets a 5-email sequence (see onboarding-sequence-design for content framework) where:

• Subject lines reference the persona's specific use case
• CTAs link to the persona's activation action (not generic dashboard)
• Use case examples match the persona's industry or role
• Skip logic exits users from the sequence when they reach persona-specific activation

Using loops-audience, segment contacts by persona_type and enroll each new contact into their matching sequence automatically.

4. Add contextual in-app messages per persona

Using intercom-in-app-messages, create nudge messages for each persona's common stall points:

• If persona = team_lead and has not invited teammates after 48 hours: show tooltip "Invite your team to get started together"
• If persona = technical_builder and has not connected an integration after 72 hours: show banner "Connect your first integration in 2 minutes"
• If persona = solo_creator and has not created a project after 24 hours: show tooltip pointing to "New project" button

Each message fires once. Dismissed messages do not return. Messages do not overlap with active product tour steps.

5. Set up per-persona feature flags

Using posthog-feature-flags, create a multi-variant feature flag onboarding-persona-v2 with one variant per persona. This flag controls:

• Which product tour starts
• Which dashboard layout the user sees on first login
• Which sample data or templates are pre-loaded
• Which help articles appear in the Intercom help center widget

Route flag assignment based on the persona_type property set in Step 2.

6. Build per-persona funnels and dashboards

Using posthog-cohorts and posthog-custom-events, build one activation funnel per persona:

signup_completed (persona=X)
  → tour_started (persona=X)
  → tour_completed (persona=X)
  → activation_reached (persona=X)

Also build a comparative dashboard showing all personas side by side: activation rate, time to activation, tour completion rate, email open rate, and email click rate. This dashboard is the primary input for identifying which persona paths need A/B testing.

7. Run per-persona A/B tests

For the worst-performing persona (lowest activation rate), use posthog-feature-flags to test variations:

• Test A: Different tour structure (3 steps vs 5 steps)
• Test B: Different first action (simpler initial milestone)
• Test C: Email-first vs tour-first (which channel drives activation better for this persona?)

Run one test at a time per persona. Require 100+ users per variant before evaluating. Use the ab-test-orchestrator drill for statistical rigor.

Output

• 5+ persona-specific onboarding paths (tours + emails + in-app messages)
• Automated persona classification assigning 95%+ of new signups without manual tagging
• Per-persona activation funnels showing comparative performance
• At least one completed A/B test per persona improving the weakest activation paths