Who This Helps
This is for you, Junior Analyst. You just saw a KPI drop and your manager wants answers by Friday. The Metrics & Dashboards Basics course gives you the exact framework to diagnose fast and ship a clean analysis.
Mini Case
Maya, a junior analyst like you, noticed the North Star Metric dropped 12% in 7 days. She had 20 numbers on her dashboard and no clue where to start. Using the course's mission on North Star Metric and Supporting Metrics & Targets, she focused on three supporting metrics: new users, activation rate, and weekly active users. She found activation rate fell 8% because a new onboarding step confused users. Her recommendation: roll back the step and add a tooltip. Her manager said yes.
Do This Now (5 Steps)
- Pick your primary metric. If you track 20 numbers, choose one North Star Metric. That's your anchor.
- List three supporting metrics. For each, write a clear definition. Example: "Activation rate = users who complete step 3 within 7 days."
- Set realistic targets. Use last month's average as a baseline. If your target was 40% and you hit 32%, you know the gap.
- Build a weekly scoreboard. Put your primary metric at the top, supporting metrics below. Add guardrails: if a metric drops below 30%, flag it.
- Diagnose in one session. Look at the flagged metric. Ask: "What changed 7 days ago?" Check release logs, campaign dates, or data pipeline issues. Write one root cause and one recommendation.
Avoid These Traps
- Don't chase every number. Focus on three supporting metrics max.
- Don't skip the definition. Vague metrics lead to vague fixes.
- Don't blame the data first. Check if a recent change caused the drop.
- Don't write a long report. One page with root cause and recommendation is enough.
- Don't forget guardrails. Without them, you'll miss the drop until Friday panic.
Your Win by Friday
By Friday, you'll have one clear root cause and one recommendation. Your manager will see a clean analysis, not a data dump. And you'll feel like a detective who solved the case. Plus, you'll have a repeatable process for next week's drop. That's a win.