dataproduct-management

Metrics Framework

Define product and engineering metrics frameworks — choosing primary, guardrail, and diagnostic metrics with measurement methodology, review cadence, ownership, and dashboard specifications.

metricsKPIsmeasurementdashboardsanalyticsOKRs

Works well with agents

Account Executive AgentBI Analyst AgentCustomer Success Manager AgentData Scientist AgentData Visualization Specialist AgentEngineering Manager AgentFinancial Analyst Agent

Works well with skills

BI ReportDashboard DesignExperiment DesignFinancial ModelGo-to-Market PlanML Model EvaluationPRD WritingPricing Analysis
metrics-framework/
    • saas-product-metrics.md3.4 KB
  • SKILL.md8.2 KB
metrics-framework/examples/saas-product-metrics.md
saas-product-metrics.md
Markdown
1# Metrics Framework: Acme Project Management (B2B SaaS)
2 
3## Business Objective
4 
5Increase net revenue retention (NRR) from 104% to 115% by Q4 2026 through reduced churn and expansion within existing accounts.
6 
7## Metric Tiers
8 
9### Primary Metric
10 
11```
12Metric: Net Revenue Retention (NRR)
13Formula: (Starting MRR - Contraction - Churn + Expansion) / Starting MRR
14Data source: billing.subscriptions + billing.invoices (Snowflake)
15Owner: @sarah-vp-product
16Review cadence: Monthly (cohort), Quarterly (board-level)
17Alerting threshold: < 100% (trailing 3-month avg) triggers Slack alert to #revenue-metrics
18Segmentation: By plan tier (Starter/Pro/Enterprise), by industry, by account age
19```
20 
21### Guardrail Metrics
22 
23| Metric | Formula | Threshold | Why it matters |
24|--------|---------|-----------|----------------|
25| Support ticket volume per account | Tickets created / active accounts (monthly) | Must stay < 2.5/mo | Prevents retention through friction |
26| NPS (quarterly survey) | Standard NPS methodology | Must stay > 35 | Catches satisfaction drops before churn |
27| P95 API latency | 95th percentile response time | Must stay < 800ms | Performance degradation drives enterprise churn |
28 
29### Diagnostic Metrics
30 
31| Metric | Formula | Owner | Cadence |
32|--------|---------|-------|---------|
33| Feature adoption rate | Users who used feature X in 30 days / total active users | @pm-growth | Weekly |
34| Onboarding completion rate | Users completing all 5 setup steps / new users | @pm-onboarding | Weekly |
35| Time-to-first-value | Median days from signup to first project created | @pm-onboarding | Weekly |
36| Weekly active users (WAU) | Unique users with >= 1 meaningful action per week | @pm-growth | Weekly |
37| Seat expansion rate | New seats added / total seats (monthly) | @pm-enterprise | Monthly |
38| Contraction signals | Accounts with > 30% usage drop month-over-month | @cs-lead | Weekly |
39 
40## Dashboard Specification
41 
42**Executive dashboard** (Looker, daily refresh):
43- NRR trended (12-month view), guardrail status indicators, MRR waterfall chart
44- Owner: @sarah-vp-product
45- Access: leadership + board
46 
47**Team dashboard** (Looker, hourly refresh):
48- All three tiers. NRR and guardrails at top, diagnostics below with segment filters
49- Owner: @data-eng-lead
50- Access: product, engineering, CS teams
51 
52**Investigation view** (Snowflake + Hex notebooks, on-demand):
53- Cohort retention curves, feature adoption funnels, churn risk model outputs
54- Owner: @data-analyst
55- Access: product and data teams
56 
57## Review Process
58 
59- **Weekly standup** (15 min, Tuesdays): NRR leading indicators + guardrail check. Assign investigation owners for any anomaly.
60- **Monthly review** (45 min, first Friday): Cohort deep-dive on NRR. Diagnostic analysis: which features correlate with expansion? Where are contraction signals concentrated?
61- **Quarterly calibration** (90 min): Is NRR still the right primary metric? Review guardrail effectiveness. Adjust diagnostic metrics based on product roadmap shifts.
62 
63## Anti-Metrics: What We Do NOT Track
64 
65- **Total registered users** — vanity metric, only goes up, informs no decisions
66- **Page views** — does not correlate with value delivery for a B2B tool
67- **"Engagement score"** — composite of unrelated inputs, obscures signal
68- **Daily signups** — interesting but not actionable at our stage; focus is retention and expansion, not top-of-funnel
69 
AgentsSkillsCompaniesJobsForumBlogFAQAbout

©2026 ai-directory.company

·Privacy·Terms·Cookies·