Pricing Health Monitor

This drill builds an always-on monitoring system for pricing-related metrics. It detects when pricing is causing churn, when usage patterns shift beyond current tier boundaries, and when revenue efficiency changes. This is the observation layer that feeds the autonomous-optimization drill at Durable level.

Prerequisites

• billing-event-streaming fundamental configured (Stripe events flowing into PostHog)
• At least 90 days of billing + usage data in PostHog
• n8n instance for scheduled monitoring

Steps

1. Build the pricing health dashboard

Using posthog-dashboards, create a "Pricing Health" dashboard with these panels:

Revenue Panels:

• ARPU trend (30-day rolling): total revenue / active customers
• ARPU by plan: breakdown per pricing tier
• Net Revenue Retention (NRR): monthly, trailing 12-month
• Expansion revenue: revenue from upgrades and usage growth
• Contraction revenue: revenue lost from downgrades

Churn Panels:

• Churn rate by plan: which plan has highest churn?
• Churn rate by usage band: are low-usage users churning more? (pricing too high for value received)
• Churn rate by tenure: is churn front-loaded (onboarding) or distributed (pricing fatigue)?
• Revenue churn vs. logo churn: losing big accounts or many small ones?

Usage-to-Revenue Panels:

• Usage per dollar spent: are customers getting more value per dollar over time? (good — means stickiness)
• Percentage of customers at plan limits: >30% means your tiers are too tight
• Overage revenue as percentage of total: >40% means the base plan is underpriced
• Usage growth rate vs. revenue growth rate: if usage grows faster than revenue, your pricing is leaking value

2. Set anomaly detection rules

Using posthog-anomaly-detection, configure alerts for:

| Metric | Alert Condition | Severity | |--------|----------------|----------| | ARPU | Drops >10% month-over-month | High | | NRR | Falls below 100% for 2 consecutive months | Critical | | Churn rate (any plan) | Exceeds 2x the 90-day average | High | | Customers at plan limit | Exceeds 40% of plan subscribers | Medium | | Overage complaints | >5 support tickets mentioning "bill" or "charge" in 7 days | High | | Expansion revenue | Drops >20% month-over-month | Medium |

3. Build the daily monitoring workflow

Using n8n-scheduling, create a workflow that runs daily at 07:00 UTC:

1. Query PostHog for all pricing health metrics (ARPU, NRR, churn by plan, usage distribution)
2. Compare each metric against its 30-day rolling average and the anomaly thresholds from Step 2
3. If any metric is anomalous: a. Log pricing_anomaly_detected event in PostHog using posthog-custom-events with properties: metric_name, current_value, baseline_value, severity, affected_plan b. Create an Attio note on the pricing project record with the anomaly details c. If severity = Critical, send a Slack alert to the pricing owner
4. If all metrics are normal, log pricing_health_check_passed event

4. Build the weekly pricing digest

Using n8n-scheduling, create a weekly workflow (Monday 09:00 UTC):

1. Aggregate the full week's pricing health data
2. Compute week-over-week changes for all metrics
3. Identify the top 3 trends (positive or negative)
4. Generate a structured report:

   Pricing Health Digest — Week of {date}
   
   ARPU: ${current} ({change}% WoW)
   NRR: {current}% ({change}pp WoW)
   Churn by plan:
     - Free: {rate}%
     - Starter: {rate}%
     - Pro: {rate}%
   
   Top trends:
   1. {trend description}
   2. {trend description}
   3. {trend description}
   
   Action items:
   - {item if anomaly detected}
   

5. Post to Slack and store in Attio

5. Track pricing sensitivity cohorts

Using posthog-cohorts, maintain dynamic cohorts that signal pricing stress:

• Pricing page visitors (active customers): Users who visited the pricing/billing page 2+ times in 7 days while on a paid plan. This correlates with considering a change.
• Limit-approaching: Users at 80%+ of a plan limit. They will either upgrade or churn.
• Overage shock: Users whose last invoice was 50%+ higher than their previous invoice (usage spike causing bill shock).
• Downgrade candidates: Users whose usage dropped below the threshold for their current plan for 2+ consecutive weeks.

Update these cohorts daily. Feed them into the churn-prevention and upgrade-prompt drills.

6. Calibrate monthly

At the end of each month:

1. Compare predicted anomalies vs. actual outcomes. Were the alerts accurate?
2. Adjust thresholds: if false positives exceed 30%, tighten thresholds. If misses exceed 20%, loosen them.
3. Check if the usage distribution has shifted enough to warrant tier boundary changes (feed back into usage-pricing-model-analysis)
4. Log calibration results as a PostHog event: pricing_monitor_calibration with precision, recall, threshold_changes

Output

• A PostHog dashboard with real-time pricing health visibility
• Daily anomaly detection with severity-based alerting
• Weekly pricing digest posted to Slack
• Dynamic cohorts tracking pricing stress signals
• Monthly calibration loop for monitoring accuracy

Triggers

Daily monitoring runs via n8n cron. Weekly digest on Mondays. Monthly calibration on the 1st. On-demand via webhook for ad-hoc checks after pricing changes.