DS-101 · Module 3

Building Your Dashboard

3 min read

You do not need a data team to build a personal dashboard. You need five metrics, a spreadsheet or simple tool, and fifteen minutes every Monday morning. That is the entire system. The weekly data review habit is more valuable than any enterprise BI platform, because the habit forces you to look at the numbers consistently — and consistency is what turns noise into pattern recognition.

  1. Pick Your Five Metrics Use the metric stack from Lesson 6. One North Star, two supporting metrics, two diagnostic metrics. If you are in sales, it might be: pipeline value, qualified meetings this week, proposals sent, average deal size, and win rate. Write them down. These are the only numbers you check weekly.
  2. Set Up a Simple Tracker A spreadsheet is fine. A Notion table is fine. The tool does not matter. What matters is that you record the number every week in the same place. Over time, the weekly snapshots become a trend line. After 8 weeks, you will start seeing patterns that daily glances never reveal.
  3. Review Every Monday for 15 Minutes Look at each metric. Compare it to last week. Compare it to the 4-week average. If something moved significantly, write one sentence explaining why. If you do not know why, that is your investigation for the week. The act of writing the explanation forces clarity.
  4. Set Anomaly Alerts Use AI or simple threshold rules to flag when a metric moves more than 15% from its 4-week average. Most tools — even spreadsheets with conditional formatting — can do this. The alert saves you from missing the one week where the number moved and you were too busy to notice.

The weekly data review habit compounds. In week one, you are just recording numbers. By week four, you notice your first pattern. By week eight, you are predicting next week's numbers before you check them — and you are right more often than you are wrong. By week twelve, you are making decisions faster because the data context is already in your head. You do not need to "pull a report." You already know.