Who This Helps
This is for junior analysts who need to move from 'the number is down' to 'here's why and what we should do.' It uses the core dashboard skills from the Metrics & Dashboards Basics course.
Mini Case
Your weekly scoreboard shows a 15% drop in user sign-ups. The team is pointing fingers at the new homepage, but your dashboard's supporting metrics tell a different story. Traffic is up 10%, but the conversion rate on the sign-up form plummeted by 25% last Wednesday. That's your real signal.
Do This Now (5 Steps)
- Open your weekly scoreboard. Don't get lost in a sea of tabs. Start with the one dashboard built for calm weekly decisions.
- Check your North Star and its guardrails. Is the drop breaking a critical threshold you set? This tells you the severity.
- Look one level down at your supporting metrics. The course teaches you to build a metric tree for this exact moment. Which supporting metric moved first or most?
- Note the exact date of the change. Pinpoint if it happened over 2 days or 7. This narrows the search for external causes.
- Write one sentence linking the cause to the main KPI. For example: 'Sign-ups dropped 15% because form conversion fell 25% after the Wednesday page update.' Boom. You have a hypothesis.
Avoid These Traps
- Don't start deep-diving into raw data before checking your high-level dashboard. You'll waste an hour.
- Don't blame 'seasonality' or 'external factors' until you rule out changes in your own supporting metrics.
- Avoid presenting three possible root causes. Your job is to diagnose the most likely one. Be a detective, not a list-maker.
- Don't forget to look at the week before the drop. Sometimes the real story is that a previous spike has just ended.
Your Win by Friday
You'll walk into your weekly sync able to say: 'Our main KPI dropped, but the data points to this specific lever. Here's my recommended next step.' You'll replace panic with a clear path forward. That's the power of a well-built dashboard—it turns noise into a clear signal. You've got this.