PM Mapped
Home / Module 5 / Building Dashboards
23
MODULE 5 · ANALYTICS IN PRACTICE · TOOL 23

Building Dashboards

The discipline of building dashboards that drive decisions rather than just display data. The core principle is a definition: a dashboard is a decision-making tool, not a data display.

▸ Try the interactive tool
SolvesDashboards nobody reads because they answer no question.
Category · Analytics in Practice Complexity · Beginner–Mid Time to apply · Per dashboard Pairs with · The Measurement Hierarchy
A WHAT IT IS

The framework

Building effective dashboards is the discipline of creating dashboards that drive decisions rather than merely display data. The central principle is a definition: a dashboard is a decision-making tool, not a data display. Every metric on it should answer a specific question someone will act on — if a metric prompts no decision, it doesn't belong.

Most dashboards fail by being data displays — crammed with every available metric, impressive-looking, and useless for actually deciding anything. The discipline reverses this: start from the decisions the dashboard's audience needs to make, then include only the metrics that inform those decisions, arranged so the important signal is immediate. A good dashboard has a clear audience, answers their specific questions, surfaces what needs attention, and ruthlessly excludes everything that's merely interesting. The test for any metric: what decision does this drive?

THE DEFINING PRINCIPLE

A dashboard is a decision-making tool, not a data display. Every metric must answer a specific question someone will act on. The test for inclusion: what decision does this metric drive? No decision → it doesn't belong.

TRY IT

Try it yourself

B WHY IT MATTERS

What it prevents

The default dashboard crams in every available metric to look comprehensive — producing an impressive wall of data that drives no decisions and hides the few signals that matter.

The shortcutWhat it costsWhat it gives you instead
Data-display dashboardsEvery metric included; nothing drives a decision.A decision-first dashboard includes only what's acted on.
Metric overloadImportant signals drowned in comprehensive clutter.Ruthless exclusion surfaces what actually matters.
No clear audienceA dashboard for everyone serves no one's decisions.A defined audience makes the dashboard answer their questions.
Vanity on the wallImpressive-looking numbers that prompt no action.The 'what decision?' test removes decoration.
C HOW TO RUN IT

Step by step

1

Define the audience and their decisions

Who uses this dashboard, and what decisions do they need to make with it? A dashboard with no specific audience and decisions becomes a data display by default.

2

Include only decision-driving metrics

For each candidate metric, ask: what decision does this drive? If the honest answer is 'none, but it's interesting,' leave it off. This ruthless exclusion is the core discipline.

3

Arrange for immediate signal

Put the most important, attention-needing information where it's seen first. A good dashboard surfaces what needs action; it doesn't make the viewer hunt.

4

Make it answer questions at a glance

The audience should get the answer to their key question quickly, not have to interpret a wall of charts. Clarity beats comprehensiveness.

5

Prune regularly

Dashboards accrete metrics over time, drifting back toward data displays. Periodically re-apply the 'what decision?' test and remove what's no longer acted on.

D IN PRACTICE

A short illustration

IN PRACTICEthe comprehensive, useless dashboard

A team built a dashboard packed with every metric they could pull — dozens of charts, impressive at a glance, comprehensive by design. In practice no one used it to decide anything: the few metrics that actually mattered were buried among the many that were merely interesting, and finding the signal took longer than just asking someone.

Rebuilding it around decisions transformed it. They started from the question 'what does this dashboard's audience need to decide?' and included only the metrics that informed those decisions — a handful, not dozens — arranged so the important signal was immediate. The comprehensive wall of data became a genuine decision tool, precisely because most of the data was removed.

The lesson: a dashboard is a decision-making tool, not a data display. The instinct to be comprehensive produces impressive, useless walls of metrics — the discipline is ruthless exclusion, keeping only what drives a decision someone will actually make.
E THE ARTIFACT

The decision-driving dashboard

The deliverable is a dashboard built from its audience's decisions outward — only decision-driving metrics, signal surfaced first.

Data display (trap)Decision tool (goal)
Every available metricOnly decision-driving metrics
Comprehensive, clutteredFocused, clear
No specific audienceBuilt for an audience's decisions
Signal buriedSignal surfaced first
F THE SO-WHAT

Why it matters

THE KEY INSIGHT

The single principle — a dashboard is a decision tool, not a data display — resolves almost every dashboard question. The test 'what decision does this metric drive?' is what keeps a dashboard useful instead of impressive.

The instinct that ruins dashboards is the urge to be comprehensive — to include every metric because each is individually interesting and leaving one out feels like a gap. But comprehensiveness is the enemy of decision-making: the more metrics on a dashboard, the harder it is to find the few that matter, until the dashboard becomes a wall of data that's consulted by no one and acted on never. The discipline of ruthless exclusion — keeping only metrics that drive a specific decision for a specific audience — feels like leaving value out, but it's what creates value, because a dashboard's worth is measured in decisions made, not metrics displayed. This also connects to the measurement hierarchy: a good dashboard shows the right level of metric for its audience's decisions, not every level at once.

G MISTAKES & LIMITS

Common mistakes

Building a data display

Comprehensive isn't useful. Include only metrics that drive decisions.

Metric overload

Burying the signal in clutter. Ruthlessly exclude the merely interesting.

No defined audience

A dashboard for everyone decides nothing. Build for a specific audience.

Never pruning

Dashboards accrete metrics and drift to displays. Re-apply the decision test regularly.

When not to use it

H CONNECTS TO

Where this sits in the toolkit

Shows the right → Measurement Hierarchy level

A dashboard should surface the metric level (Tool 02) its audience decides with.

A category in → the Analytics Stack

BI/dashboards is one of the stack's categories (Tool 21).

Often built on → SQL

Dashboard metrics frequently come from SQL queries (Tool 22).

Guards against → Vanity Metrics

The 'what decision?' test excludes the vanity metrics of Tool 27.

I WORKING WITH AI

How AI changes this in practice

AI speeds up dashboard building — drafting queries, suggesting visualisations — but the discipline of what to include is still yours.

  • Generate the underlying queries: describe the metric and let AI draft the SQL behind each tile (verify it).
  • Suggest visualisations: ask which chart type best fits a given metric and why.
  • Draft summaries: have it write the plain-English 'what this dashboard tells you' caption.

The judgment that stays yours: A dashboard is a decision tool, not a data display — and deciding which metrics earn a place is judgment AI can't do for you. It'll happily help you build a cluttered dashboard of vanity metrics. Use it for the build, not the editorial call.

Go deeper → full AI guide with examples & a copy-paste template
TRY IT YOURSELF

Apply the decision test

Take a dashboard you've seen (or imagine a typical one). For each metric, ask honestly: what decision does this drive? Mark the ones where the answer is 'none, just interesting.'

How many would survive ruthless exclusion — and would removing the rest make the dashboard more useful or less?

If cutting the no-decision metrics makes the dashboard clearer and more actionable, you've proven the principle: a dashboard's value is in decisions driven, not metrics displayed.