Dashboard Design Principles for Data Analysts — Layout, KPIs, Colour & Stakeholder Dashboards
A technically correct dashboard that nobody uses is a failed dashboard. This chapter teaches the design principles that make dashboards decision-ready — covering KPI hierarchy, layout grids, colour systems, chart selection rules, and the most common mistakes analysts make when building for stakeholders.
⏱ 20 min read📅 September 2026✍ EVIKA ACADEMY, Noida
Why Dashboard Design Is an Analyst Skill, Not a Designer Skill
Most analysts treat dashboard design as an afterthought — get the data right, then drag charts onto a canvas. The result is dashboards that answer every question except the one the stakeholder actually has. Good dashboard design is not about aesthetics; it is about reducing the cognitive load required to extract a decision from data.
🎯
One dashboard, one purpose
A dashboard built for a CEO and one built for operations will look completely different, even with the same data.
⚡
5-second rule
A stakeholder should extract the key insight within 5 seconds of opening the dashboard. If they need to hunt, it has failed.
🔄
Designed to be updated
Dashboards that require manual refresh get abandoned. Design for automatic or one-click refresh from day one.
📱
Mobile aware
Executives increasingly check dashboards on phones. A 4-column grid that works on a monitor collapses into noise on a screen.
The KPI Hierarchy — What Goes Where
Before opening Power BI or Excel, write down your KPIs in three tiers. This hierarchy determines the visual weight and position of every element on your dashboard.
Tier 1 — North Star Metric
The single number that defines whether the business is winning or losing this period. Goes top-left or top-centre in the largest font. Maximum one per dashboard.
Examples: Monthly Revenue, Total Orders, Active Users, Gross Margin
Tier 2 — Supporting KPIs
The 3–5 metrics that explain the North Star. Go in a card row immediately below or beside Tier 1. Equal visual weight, smaller than Tier 1.
Examples: New vs Returning Revenue, Avg Order Value, Conversion Rate, Churn Rate
Tier 3 — Diagnostic Metrics
Breakdowns and trends that explain why Tier 2 moved. Go in the lower half as charts and tables. Only appear when Tier 2 flags a problem.
Examples: Revenue by Channel, Top 10 Products, City-wise Breakdown, Weekly Trend
Rule:If a stakeholder has to scroll to see the Tier 1 metric, your layout hierarchy is wrong. The most important number must be visible without any interaction.
Layout Grid — The F-Pattern and Z-Pattern
Eye-tracking research consistently shows that people scan dashboards in an F-pattern (left to right, then down the left side) or Z-pattern (top-left to top-right, then diagonal to bottom-left, then bottom-right). Design your layout to place the highest-priority content where eyes land first.
RECOMMENDED DASHBOARD GRID (Power BI / Excel)
₹2.4Cr
NORTH STAR — Monthly Revenue
18.4%
Conversion Rate
₹1,840
Avg Order Value
📈 Revenue Trend (line chart — 13 weeks)
🏆 Top 5 Products (horizontal bar)
📊 By Channel (donut)
🗺 By City (map)
📋 Detail table
👁 Eyes land top-left → read right across KPI row → drop to charts → bottom-right detail
✅ Left-align the most important metric
F-pattern scanning means the left column gets the most attention. Put your North Star there.
✅ Keep equal-weight KPIs in a horizontal card row
Cards at the same size signal equal importance. Varying size creates accidental hierarchy.
Dense dashboards feel overwhelming. White space is not wasted space — it is breathing room that makes numbers readable.
Colour System — Semantic Colours That Communicate Instantly
Colour in a dashboard is not decoration — it is communication. Every colour choice should answer the question: does this colour help the user make a faster, better decision?
SEMANTIC COLOUR SYSTEM
#0e7490
Primary brand
Titles, key metric values, primary chart series
#16a34a
Positive / on track
Revenue up, target met, green traffic light
#d97706
Warning / watch
Approaching threshold, slight miss, amber light
#dc2626
Negative / alert
Target missed, churn spike, red traffic light
#f1f5f9
Neutral / background
Card backgrounds, alternating table rows
#0f172a
Text primary
All heading text and metric values
#64748b
Text secondary
Labels, subtitles, axis text, footnotes
3 Colour Rules Every Analyst Must Follow
✅ Never use more than 3–4 colours in a single chart
More than 4 colours forces the viewer to constantly refer to the legend. Use shades of one colour for the same category (e.g. dark blue = this year, light blue = last year) rather than introducing new hues.
✅ Red means bad, green means good — always
Do not use red for the top-performing bar just because it looks vivid. Stakeholders decode colour meaning in milliseconds. Violating the red/green convention causes real confusion and erodes trust.
✅ Ensure 4.5:1 contrast ratio for all text on coloured backgrounds
White text on yellow fails accessibility standards and is unreadable in bright light. Use a contrast checker tool. In Power BI: always test your theme in a well-lit room before sharing.
Chart Selection for Dashboards — Quick Reference
Dashboard charts must be readable at a glance from arm's length. Charts that work well in a standalone report often fail on a dashboard because they are too complex to decode at reduced size.
Use case
Best choice
Why it works on dashboards
Avoid
Single KPI vs target
Card with trend arrow + RAG colour
Instant read — no chart required
Gauge/speedometer
Trend over time
Line chart (thin line, minimal gridlines)
Direction readable even at 200px wide
Area chart with multiple series
Compare 3–6 categories
Horizontal bar, sorted descending
Labels readable, ranking obvious
Pie chart, 3D bar
Part-to-whole (≤4 parts)
Donut with centre label showing total
Small footprint, reads well at glance
Stacked bar across many periods
Two-metric correlation
Scatter plot (use only in detail section)
Too complex for top of dashboard
—
Geographic data
Filled map or bubble map
Spatial pattern immediately obvious
Table of city names
Funnel / conversion
Horizontal funnel bar
Drop-off between stages instantly visible
Line chart (implies time not steps)
Anatomy of a Well-Designed KPI Card
KPI cards are the most-read element of any executive dashboard. Every element of a card has a purpose — remove what is not essential.
Monthly Revenue
₹2.4Cr
+18% vs last month
Target: ₹2.2Cr ✅
Conversion Rate
18.4%
-1.2% vs last month
Target: 20% ⚠️
Avg Order Value
₹1,840
+₹120 vs last month
Target: ₹1,750 ✅
Metric label (top)
ALL CAPS, small font, muted colour — this sets context without competing with the value
Primary value (large)
The biggest text on the card. No decimal places unless precision matters. Use abbreviations: Cr, L, K.
Change vs prior period
Always show direction (▲ ▼ or + / -) with colour. MoM is most common; YoY when seasonality matters.
Target context (bottom)
One line: "Target: X ✅" or "Behind by Y ⚠️". Stakeholders need to know if the number is good or bad.
8 Dashboard Design Mistakes Analysts Make
01
❌ Putting everything on one page
Fix: A dashboard that tries to show everything shows nothing. Use tabs or drill-through pages: Overview → Category → Detail.
02
❌ Using the default theme / colours
Fix: Default Power BI blue or Excel grey communicates zero effort. Spend 20 minutes setting a custom colour theme that matches your brand or client.
03
❌ No date/time filter visible
Fix: Every dashboard needs a clear, visible date range selector at the top. Stakeholders need to know what period they are looking at without reading a footnote.
04
❌ Showing too many decimal places
Fix: "₹2,43,18,492.37" is harder to read than "₹2.43Cr". Round aggressively in dashboards — precision lives in exported reports.
05
❌ Charts without titles
Fix: Every chart needs a headline title (the insight, not the subject). A chart titled "Revenue" tells you nothing. "Revenue accelerating — Q3 is 26% ahead of Q2" tells you everything.
06
❌ Mixing time granularities
Fix: Do not put daily revenue next to monthly churn on the same trend chart. Align granularity or use separate charts with clear time-period labels.
07
❌ No mobile testing
Fix: In Power BI, use the Phone layout view. In Excel, zoom to 75% and check if the dashboard is still readable. Executives will check on their phones.
08
❌ Sharing before getting feedback
Fix: Before releasing to stakeholders, show a draft to one person who will use it daily. Their first 60 seconds of confusion will tell you everything you need to fix.
Our Power BI and Excel courses in Noida Sector 51 teach dashboard design from scratch — KPI hierarchy, colour systems, interactivity and stakeholder presentation. Online & offline classes available.
The key principles are: one dashboard, one purpose; KPI hierarchy (North Star → Supporting → Diagnostic); F-pattern layout so important metrics are top-left; semantic colours (green = good, red = bad); maximum 3–4 colours per chart; card titles as insights not topics; and always showing date range context.
How many KPIs should a dashboard have?
A well-designed dashboard has 1 North Star metric, 3–5 supporting KPIs, and 3–6 diagnostic charts. More than 7–8 total metrics on a single page overwhelms stakeholders. Use tabs or drill-through pages for additional detail.
Should I use Power BI or Excel for dashboards?
Use Power BI for recurring operational dashboards shared with multiple stakeholders, large datasets, and when you need real-time data or mobile access. Use Excel for ad hoc analysis, one-time reports, and when your audience is primarily a finance or operations team that works in Excel daily.
What colours should I use in a business dashboard?
Use a semantic colour system: one brand/primary colour for key metrics, green for positive/on-target, amber for warning, red for negative/alert, and neutral grey for backgrounds. Limit to 3–4 colours total per chart. Always ensure 4.5:1 contrast ratio for text on coloured backgrounds.
Where can I learn dashboard design in Noida?
EVIKA ACADEMY in Noida Sector 51 teaches Power BI and Excel dashboard design as part of its data analytics courses. Topics include KPI hierarchy, layout grids, DAX measures, colour systems, and stakeholder presentation. Online and offline classes are available. WhatsApp +91-8081035456 to book a free demo.