Metrics Dashboard

Documentation Index

Fetch the complete documentation index at: https://docs.getcore.me/llms.txt

Use this file to discover all available pages before exploring further.

Goal: Define the structure and content of a metrics dashboard that surfaces key business and product health indicators at a glance, aligned to strategy and actionable for stakeholders.

​

Tools Required

This skill runs using CORE memory only. No integrations required.

​

Step 1: Identify Dashboard Audience and Cadence

Ask:

• Who is this dashboard for? (CEO, product team, marketing, finance, engineering)
• How often will they check it? (daily, weekly, monthly)
• What decisions does it inform? (hiring, feature prioritization, spend allocation, escalations)
• What’s the current pain? (missing context, spreadsheet chaos, slow updates, too many dashboards)

​

Step 2: Map to Strategic Goals

Ask the user to list:

• 1-3 company OKRs — What are you trying to achieve?
• Financial metrics — How do you make money? (MRR, LTV, COGS, margin, unit economics)
• Customer metrics — How fast are you growing and retaining? (users, customers, signups, churn, NPS)
• Product metrics — Is the product healthy? (engagement, feature adoption, incident rate)

For each, identify: What metric proves progress? What’s the target? What’s the baseline?

​

Step 3: Define Metric Tiers

Organize metrics into levels of detail:

• Tier 1 (Dashboard headline) — 1-3 top-level metrics the CEO sees (e.g., MRR, growth rate, NPS)
• Tier 2 (Department view) — Metrics each team owns (e.g., Sales: pipeline, conversion; Eng: deploy frequency, incident rate)
• Tier 3 (Deep dive) — Operational metrics people drill into (feature usage, cohort retention, CSAT by segment)

​

Step 4: Choose Chart Types for Each Metric

For each metric, decide:

• Trend (line chart) — Is it going up or down over time? (e.g., MRR over 12 months)
• Breakdown (pie/stacked bar) — How is this divided? (e.g., revenue by segment, signups by channel)
• Comparison (bar chart) — How do we compare to target or last period? (e.g., actual vs. plan)
• Status (big number + context) — What’s the headline with variance? (e.g., “1,450 active users (+12% vs. last week)“)

​

Step 5: Design the Layout

Sketch the dashboard structure:

• Top row — 3-4 headline metrics with current value + trend
• Middle section — Drill-down by department (Product, Sales, Eng, Marketing)
• Bottom section — Alerts or anomalies (“Revenue down 8% WoW” “New feature adoption 5%—below target”)

Map out which metrics live where. Keep above-the-fold data as signal, detailed breakdowns in drill-down sections.

​

Step 6: Define Update Frequency and Data Sources

For each metric:

• Source — Where does the data come from? (product DB, payment processor, survey tool, manual input)
• Refresh rate — Daily? Weekly? Real-time? (affects tooling choice)
• Owner — Who keeps it current? (eng, analyst, product ops)
• Calculation — What’s the exact formula? (to avoid ambiguity)

​

Step 7: Present the Dashboard Specification

---

Metrics Dashboard: [Product/Team Name] Audience: [Role/Team], Updated [Frequency] Tier 1: Headlines Tier 2: Department Breakdowns Product Team

• Active users: [Value] (+X% vs. last week)
• Engagement (sessions/user/week): [Value] (target: [Y])
• Feature adoption (% using [Feature]): [Value] (target: [Y])
• Churn rate (monthly): [Value]% (target: [Y]%)

Sales & Marketing

• Monthly signups: [Value] (+X% vs. last month)
• Paid conversion rate: [Value]% (target: [Y]%)
• CAC (customer acquisition cost): $[Value] (target:$[Y])
• Sales pipeline: $[Value] (X deals in [stages])

Engineering

• Deploy frequency (per week): [Value]
• Incident rate: [X] incidents (target: [Y] per month)
• Build success rate: [X]% of deploys incident-free
• Technical debt score: [Value] (track quarterly)

Tier 3: Deep Dives (linked/expandable)

• [Metric 1] by customer segment
• [Metric 2] by acquisition channel
• [Metric 3] cohort analysis (retention by signup month)

Alerts & Anomalies

• 🚨 [Alert]: Revenue down [X]% vs. forecast — [Possible cause] — [Recommended action]
• ⚡ [Alert]: [Feature adoption] below target — [Possible cause] — [Recommended action]

---

​

Edge Cases

• Too many metrics (>20): Ask: “If you could only see 5 things on this dashboard, which would they be?” Start minimal; add detail later.
• Metric is hard to calculate: Flag the effort. Ask: “Is this worth engineering time to automate, or should we track it manually for now?” Default to pragmatism.
• Metrics point in different directions: (e.g., revenue up, churn up). This is real. Surface the contradiction. Ask: “Why might this be happening?” Investigate, don’t hide.
• Data sources are fragmented: Map integrations needed. Ask: “Should we build a unified warehouse first, or accept data pulls from multiple sources?” Decide on tool constraints early.
• Stakeholder wants to track everything: Redirect: “I hear you. Let’s pick 3 metrics this quarter, then add more next quarter as we get discipline around what moves the needle.”