Upgrade Prompt Health Monitor

This drill builds the measurement and alerting layer for in-app upgrade prompts. It tracks the full prompt-to-upgrade funnel per trigger type (limit proximity, feature gate, growth signal), detects performance degradation, and surfaces which prompt variants and segments are converting or failing.

Input

• Upgrade prompts deployed via the upgrade-prompt drill (at least 1 trigger type live)
• PostHog tracking upgrade_prompt_shown, upgrade_prompt_clicked, upgrade_started, upgrade_completed events
• n8n instance for scheduled monitoring
• Attio CRM with expansion deal records

Steps

1. Build the upgrade prompt funnel

Using the posthog-funnels fundamental, create a funnel for each trigger type:

upgrade_prompt_shown (trigger_type = "limit_proximity")
  -> upgrade_prompt_clicked
    -> upgrade_started
      -> upgrade_completed

Create separate funnels for:

• Limit proximity: user at 80%+ of a plan limit
• Feature gate: user attempted a gated feature
• Growth signal: usage volume or team size grew rapidly
• Time-based: sustained free usage beyond 30 days

Set conversion window to 7 days (most upgrades happen within 48 hours of prompt, but allow a full week for email follow-up conversions).

Break down each funnel by:

• plan_current (free, starter, pro) to see which plans convert
• prompt_surface (in-app modal, banner, tooltip, email) to see which delivery method works
• prompt_variant (for A/B test tracking)

2. Build the upgrade prompt dashboard

Using posthog-dashboards, create a dashboard called "Upgrade Prompts — Health":

| Panel | Type | What it shows | |-------|------|---------------| | Prompts shown this week | Trend | Daily count of upgrade_prompt_shown, broken down by trigger type | | Prompt CTR by trigger | Bar chart | Click rate per trigger type, current week vs previous 4-week average | | Upgrade funnel by trigger | Funnel | Full funnel per trigger type (shown -> clicked -> started -> completed) | | Revenue from prompts | Trend | Sum of upgrade_completed.revenue_delta per day | | Prompt fatigue signal | Trend | Ratio of upgrade_prompt_dismissed to upgrade_prompt_shown over time; rising ratio = fatigue | | Segment performance | Table | CTR and conversion by user segment (plan, company size, usage tier) | | Surface comparison | Bar chart | Conversion rate by prompt surface (in-app vs email vs tooltip) |

Set dashboard subscription to deliver Monday 09:00 to the growth team.

3. Create performance cohorts

Using posthog-cohorts, create dynamic cohorts:

• prompt-converted: Users who completed an upgrade within 7 days of seeing a prompt. Use for revenue attribution.
• prompt-fatigued: Users who dismissed 3+ prompts in 14 days without clicking. Suppress further prompts for these users.
• prompt-high-intent: Users who clicked a prompt but did not complete the upgrade. Route to a follow-up email or in-app help.
• prompt-never-shown: Active users on a free or lower plan who have never seen an upgrade prompt. Potential untapped audience.

4. Build degradation detection

Using posthog-anomaly-detection and n8n-scheduling, create a daily workflow that:

1. Queries PostHog for each trigger type's CTR over the last 7 days
2. Compares against the 4-week rolling average CTR for that trigger type
3. Flags any trigger type where CTR dropped more than 20% from its rolling average
4. Flags if the prompt-fatigued cohort grew more than 15% week-over-week

When a flag fires, the n8n workflow:

• Creates a note on the Attio campaign record for this play using attio-reporting
• Sends a Slack notification: "Upgrade prompt CTR for [trigger_type] dropped [X]% vs 4-week avg. Prompt fatigue cohort: [size]."
• If prompt-fatigued cohort exceeds 10% of active free users, recommend pausing prompts for that segment and switching to email-only nudges

5. Track prompt suppression and frequency

Using posthog-custom-events, log suppression events to prevent over-prompting:

posthog.capture('upgrade_prompt_suppressed', {
  reason: 'fatigue_threshold',  // or 'cooldown_period', 'segment_exclusion'
  trigger_type: 'limit_proximity',
  user_prompt_count_30d: promptCount,
  days_since_last_prompt: daysSinceLast
});

Monitor the suppression rate. If more than 30% of eligible prompt impressions are being suppressed, the prompt strategy needs redesign (fewer, higher-quality triggers rather than broad coverage).

6. Measure revenue attribution

Using posthog-custom-events, tag every upgrade completion with its prompt attribution:

posthog.capture('upgrade_completed', {
  trigger_type: 'feature_gate',
  prompt_surface: 'in_app_modal',
  prompt_variant: 'v2_social_proof',
  revenue_delta_monthly: mrr_increase,
  days_from_prompt_to_upgrade: daysDiff,
  attributed_to_prompt: wasPromptShownWithin7Days
});

Build a PostHog insight: monthly recurring revenue directly attributed to upgrade prompts, broken down by trigger type. This is the single number that justifies the play's existence.

Output

• Upgrade prompt funnel per trigger type in PostHog
• Health dashboard with 7 panels and weekly delivery
• 4 dynamic cohorts for segmentation and suppression
• Daily degradation detection with Slack + Attio alerting
• Prompt suppression tracking to prevent fatigue
• Revenue attribution per trigger type

Triggers

Dashboard reviewed weekly. Degradation detection runs daily via n8n cron at 08:00 UTC. Revenue attribution insight updated continuously via PostHog. Re-run full setup when adding new trigger types or prompt surfaces.