KPI & Metrics Prompt Templates

**KPI: Customer Health Score**

**Definition**: A composite score (0-100) predicting a customer's likelihood to renew based on engagement and sentiment signals.

**Formula**:
```
Health Score = (Usage Score × 0.4) + (Engagement Score × 0.3) + (Support Score × 0.2) + (Sentiment Score × 0.1)

Where:
- Usage Score (0-100): Based on login frequency and feature adoption vs. their plan tier's typical usage
- Engagement Score (0-100): Based on stakeholder breadth (# of active users / licensed seats)
- Support Score (0-100): Inverse of ticket volume and severity, adjusted for company size
- Sentiment Score (0-100): Latest NPS response, normalized
```

**Target-setting approach**:
- Red (0-40): High churn risk, immediate intervention needed
- Yellow (41-70): Moderate risk, proactive outreach recommended
- Green (71-100): Healthy, maintain relationship

Calibrate thresholds by back-testing against actual churn data. Adjust weights quarterly based on which components best predict churn.

**Potential pitfalls**:
1. Over-indexing on usage: Some customers are light users but happy. Check if usage patterns differ by customer segment.
2. Lagging indicators: NPS and tickets trail problems. Add leading indicators like declining usage trends.
3. Gaming: CS team might focus on green accounts (easy) instead of red accounts (impactful). Track 'saves' as a complementary metric.

**Complementary metric**: Churn save rate (% of red-flagged accounts that renew after intervention). Ensures the health score is actionable, not just predictive.

**Metric Audit: Content Marketing**

**Missing (Critical):**
- **Content-attributed signups**: You're measuring inputs (content) and vanity outputs (views) but not the actual goal (signups). Add UTM tracking and conversion tracking to tie content to trials.
- **Content-to-MQL rate**: What % of blog readers become marketing qualified leads?

**Vanity metrics:**
- **Page views**: High views on irrelevant topics don't help. A post that gets 100K views from people who'll never buy is worse than 1K views from your ICP.
- **Social shares**: Feels good but rarely correlates with business outcomes. People share entertainment, not purchase intent.

**Reframe existing metrics:**
- **Time on page** → Good, but segment it. Time on page for high-intent content (pricing comparison, how-to guides) matters more than listicles.
- **Posts per month** → This incentivizes quantity over quality. Replace with 'posts meeting quality threshold' or pair with a quality metric.

**Conflicting incentives:**
- Measuring 'posts published' encourages the team to write more, but if more posts = more low-quality content = lower average conversion rate, you're going backward.

**Recommended metric stack:**
1. Content-attributed trial signups (north star)
2. High-intent page views (views on bottom-funnel content)
3. Email-to-trial conversion rate
4. Content engagement rate (scroll depth, not just time)

**Remove or de-emphasize:**
- Social shares (unless social is a real channel for you)
- Raw page views (replace with segmented views)

**Metrics Hierarchy: Churn Reduction**

**Lagging (Outcome) - Quarterly/Annual view:**
- Annual churn rate (goal: <5%)
- Net Revenue Retention (accounts for expansion)
- Lost revenue from churn ($)

**Lagging (Intermediate) - Monthly view:**
- Monthly churn rate
- Churn by segment (helps identify problem areas)
- Win-back rate (% of churned customers who return)

**Leading (Predictive) - Weekly view - Your 90-day early warning:**
- Health score distribution (% in red/yellow/green)
- Usage trend slope (are customers using more or less over past 30 days?)
- Support ticket sentiment trend
- Renewal calls scheduled vs. renewals due in 90 days

**Leading (Input/Activity) - Daily view:**
- Customer Success outreach volume to at-risk accounts
- Feature adoption for sticky features
- Time to resolve critical tickets
- QBR completion rate

**How to use this hierarchy:**
1. Daily: Track inputs to ensure CS team is doing the right activities
2. Weekly: Review leading indicators to spot accounts heading toward churn
3. Monthly: Check if leading indicator improvements are flowing to lagging metrics
4. Quarterly: Evaluate if you're hitting the churn target

**Key insight**: If leading indicators improve but lagging indicators don't follow, your leading indicators aren't actually leading indicators. Revisit and adjust.

**Suggested alerts:**
- Any account drops 20+ health score points in a week
- Usage for any account drops below 50% of their historical average
- Any Enterprise account goes 30 days without login

**Counter-metrics for Resolution Time:**

1. **Reopen rate**: % of tickets reopened within 7 days of closing
 - Why: Directly catches premature closes. If someone closes without solving, the customer reopens.
 - Target: Keep below 5%
 - Red flag: Rep with fast resolution AND high reopens

2. **Customer satisfaction (CSAT) per ticket**: Survey score on closed tickets
 - Why: Even if not reopened, customer might be dissatisfied with rushed resolution
 - Target: Above 4.2/5
 - Red flag: CSAT dropping as resolution time improves

3. **First contact resolution rate**: % of tickets resolved without escalation or follow-up
 - Why: Rewards doing it right the first time, not just fast
 - Target: Above 70%
 - Red flag: FCR dropping while resolution time drops

4. **Escalation rate**: % of tickets escalated to senior support or engineering
 - Why: Reps might close tickets by telling customers 'this isn't supported' when it actually is
 - Target: Below 10%
 - Red flag: Sudden drop in escalations alongside drop in resolution time

**Recommended approach:**
Create a composite score: Speed (resolution time) accounts for 40%, Quality (CSAT + reopen rate) accounts for 60%. This lets you improve speed without sacrificing quality.

**Dashboard view for managers:**
| Rep | Avg Resolution | Reopen % | CSAT | Quality Score | Overall |
|-----|---------------|----------|------|---------------|---------|