Marketing DashboardsFewer charts, faster decisions

Why most dashboards are ignored#

Ask any marketing team how many dashboards it has and the answer is usually 'too many'. Ask how many are opened weekly and the answer shrinks to one or two. The rest were built with enthusiasm, filled with every metric the connectors could reach, and abandoned once nobody could tell what they were for.

The cause is nearly always the same: the dashboard was designed from the data outwards. Someone connected GA4, the ad platforms and the CRM, then arranged the available numbers attractively. A useful dashboard is designed from the decisions inwards. It answers 'are we on track, and if not, where should we look?' in under a minute.

If a chart has never prompted a question in a meeting, it is wallpaper.

Step 1: write the questions before touching a tool#

Sit with the people who will use the dashboard and list the questions they need answered on a regular rhythm. A chief executive might ask: are we generating enough qualified pipeline to hit the quarter, and is acquisition cost moving? A performance marketer asks: which campaigns are above or below target cost per qualified lead this week?

Different questions mean different dashboards. Resist the urge to build one view for everyone. A leadership dashboard and a channel manager's dashboard share data but not design; trying to serve both produces a screen that serves neither. Map each question to the metric that answers it, using your agreed marketing KPIs.

Write the questions down and keep them.

Step 2: design three layers#

The outcome layer should fit on one screen without scrolling and be readable by anyone in the business. The driver layer explains movement in the outcomes: if qualified leads fell, was it volume of traffic, conversion on the landing page or qualification rate in sales? The diagnostic layer is where specialists dig. It can be detailed because only people investigating a specific question go there.

This structure also solves the political problem of dashboards. Everyone's favourite metric can have a home, but only the outcome layer gets the prime position, and that layer is agreed with leadership.

Step 3: give every number a comparison#

A number on its own is nearly meaningless. Is a cost per lead of ₹1,200 good? It depends on the target, last month, the same month last year and what a lead is worth. Every outcome and driver metric should sit beside at least one comparison.

Step 4: get the data right before the design#

A beautifully designed dashboard on untrusted data does more harm than no dashboard, because it lends authority to wrong numbers. Before building, confirm that definitions are written down, that UTM parameters are consistent, that GA4 key events match CRM outcomes within an explainable margin, and that cost data from each platform arrives in the same currency and time zone.

For anything beyond a simple dashboard, bring data into a single store first (a spreadsheet for small teams, a warehouse such as BigQuery for larger ones), join it there, and point the dashboard at the joined table. Connecting a dashboard tool directly to five live sources and blending them on screen is fragile, slow and hard to audit. Our Looker Studio guide covers this in more detail.

Step 5: design for reading, not decorating#

• Put the most important number top left. That is where most readers look first in left-to-right layouts.
• Use the simplest chart that works. Line charts for trends, bars for comparisons, a single number with a comparison for key metrics. Avoid pies with many slices and gauges.
• Use colour for meaning only. Reserve red and green (or a colour-blind-safe pair) for off-target and on-target, so colour carries a signal.
• Label in plain language. 'Qualified leads' rather than an event name; units shown; definitions available on hover or in a footnote.
• Show freshness. A last-updated timestamp stops people making decisions on stale data.

One more rule matters more than any visual guideline: never mix definitions on the same screen. If leads from GA4 sit beside leads from the CRM, readers will compare them and draw conclusions from a gap that is only a difference in counting. Choose one source per metric, label it, and keep reconciliation in a separate diagnostic view where analysts can explain it.

Mobile viewing deserves thought too. Many senior readers will open a shared dashboard link on a phone between meetings. Test the outcome layer at phone width; if the key numbers are unreadable there, the dashboard will be judged on a screenshot someone forwards on WhatsApp instead.

Step 6: build the ritual, then prune#

Dashboards survive when they are part of a recurring conversation. Attach the leadership dashboard to a weekly or fortnightly meeting with a fixed agenda: outcomes against target, the drivers that explain any gap, decisions and owners. If a dashboard has no meeting, it will drift into irrelevance.

Review usage every quarter. Remove charts nobody has asked about, merge dashboards that overlap and retire views built for projects that have ended. Pruning is the most neglected dashboard skill and the one that keeps the rest trusted.

For presenting dashboard insight to directors, see our guide to marketing reporting for the board; for the metrics to put at the very top, see the north star metric.

Published by Fabulous.Media, a network of specialist marketing agencies. Updated 9 October 2026. Platform features change often; check current official documentation before acting on platform-specific detail.