Who this is for
Business owners and operations managers planning a dashboard or reporting project.
Question this answers
How do I plan a dashboard that people actually use and that drives better decisions?
What you'll leave with
- Why most dashboards end up unused (and how to avoid it)
- How to choose metrics that matter
- Layout and design principles for business dashboards
- Data pipeline planning essentials
Why most dashboards fail
Most dashboard projects start with enthusiasm and end with a screen that nobody looks at. The failure isn't technical — it's conceptual. The dashboard was built around "what data we have" instead of "what decisions we need to make."
A dashboard crammed with 30 metrics is just a wall of numbers. A dashboard with 6 well-chosen metrics that answer specific questions is a decision-making tool.
Start with decisions, not data
Before selecting any metrics, answer these questions:
- Who will use this dashboard? The CEO needs different information from the operations manager.
- What decisions do they need to make? "Should we hire more staff?" "Which product line needs attention?" "Are we on track for quarterly targets?"
- How often do they need it? Real-time for operations. Daily for management. Weekly or monthly for strategy.
- What action should the dashboard trigger? Every metric should have a response: "If this number drops below X, we do Y."
Choosing the right metrics
Good dashboard metrics share three qualities:
- Actionable: Seeing the number triggers a specific response
- Comparable: Can be compared over time, against targets, or between segments
- Trustworthy: The data behind it is reliable and understood
Leading indicators (pipeline value, website traffic, customer enquiries) tell you what's coming. Lagging indicators (revenue, churn, profit) tell you what already happened. The best dashboards include both.
Five to eight metrics is the sweet spot. If you can't fit it on one screen without scrolling, you have too many.
Layout and design principles
- Top-left priority: The most important metric goes top-left (where eyes land first)
- Visual hierarchy: Big numbers for critical KPIs, smaller charts for supporting detail
- Colour with purpose: Green for on-track, red for attention needed — not decoration
- Context always: Numbers without context are meaningless. Show targets, trends, or comparisons alongside every metric
- One screen: If it doesn't fit on one screen, it's not a dashboard — it's a report
Setting up data sources
The dashboard is only as good as its data. Plan the pipeline:
- Source systems: Which systems contain the data? (CRM, accounting, website analytics, POS, ERP)
- Data extraction: How does data get out? APIs, database queries, file exports.
- Transformation: What needs to happen to raw data before it's dashboard-ready? Calculations, aggregations, deduplication.
- Storage: Where does the processed data live? A data warehouse, a reporting database, or directly from source.
- Refresh frequency: How often does the dashboard update? Real-time, hourly, daily?
Dashboard planning checklist
Before building your dashboard
- Users and their decision-making needs are documented
- Metrics pass the "action test" — each one triggers a response if it changes
- Metric count is 5-8 (not 20+)
- Data sources are identified and accessible
- Refresh frequency matches the decision-making cadence
- Targets or benchmarks exist for each metric
- The least technical user has been consulted on layout and usability
Key takeaways
- The best dashboards answer specific questions — not display every data point available
- Start with "what decisions does this dashboard need to support?" not "what data do we have?"
- Five to eight metrics per dashboard is the sweet spot; more than that dilutes attention
- A dashboard is only as good as its data pipeline — plan for data freshness, accuracy, and reliability
- Build for the least technical user who needs it, not for the data analyst who builds it