Designing with Intent: The Power of Visual Cues in BI Dashboards

It's been a while since I wrote about Power BI design. Recently, I've been deep in agentic development, building AI tools that create reports automatically.

And I keep hitting the same problem.

My agent creates functional reports. The data is correct. The calculations work. But when I ask it to design a page, it tries to cram everything I ask for onto a single screen. No hierarchy. No breathing room. Just... Everything, all at once.

It made me realise something: my agent doesn't understand Intent.

It doesn't know what matters most. It doesn't guide the user's eye. It just displays data.

Which brought me back to this blog post I wrote ages ago but never published. Because whilst I've been teaching agents to build reports, I forgot to teach them (and maybe remind myself) about the fundamentals of why design works.

Here's the question: Have you ever opened a dashboard and felt... Lost?

Not because the data was complicated. Not because the calculations were wrong. But because nothing told you where to start, what mattered, or what you could do.

That's what happens when visual cues are missing.

We obsess over DAX performance. We debate star schema versus snowflake. We argue about colour palettes. But we often overlook the small design choices that help users navigate our work.

Those small choices? They're called Visual cues. And they're doing most of the heavy lifting in your dashboard, whether you realise it or not.

What Are Visual Cues?

Visual cues are design elements that guide attention, clarify information, and suggest actions. They're not decorative. They're functional communication tools.

Think about:

• Colour showing importance or grouping categories

• Font weight creating hierarchy

• Icons adding instant meaning

• Spacing showing relationships

• Shadows highlighting key elements

  
  

These cues create order. They help users prioritise and process data without thinking about it.

To understand why these work, we need to talk about Don Norman.

Affordances and Signifiers: Design Theory That Matters

Don Norman wrote The Design of Everyday Things, which should be required reading for anyone building dashboards. He introduced two concepts that explain why visual cues work:

Affordances

What an object allows you to do. A button affords clicking. A slider affords dragging.

Signifiers

The visual hints that show possible actions. A shadow beneath a button. A hand cursor on hover. A bright highlight.

In dashboards:

• A hover effect on a card → Signifier that it's interactive

• A button with rounded corners and shadow → Affordance for clicking

• Consistent spacing around filters → Signifier of a functional group

Theory is great. But let's see this in practise with some Power BI examples.

Visual Cues in Data Journalism

I'm obsessed with Financial Times data journalism. Not just their reporting, but how openly they share their design thinking.

Look at The Discourse. Instead of walls of text:

• Stylised quotes break up the narrative

• Small illustrations add context

• Audio snippets create variety

• Bold subheadings guide the eye

Each element is a signifier, helping readers navigate a nonlinear story. It's elegant. It's purposeful.

Your dashboards can do the same.

Power BI Tools for Visual Cues

Power BI gives you most of the things we need. You just need to use it intentionally.

Conditional Formatting

Highlight what matters:

• Format backgrounds, font colours, icons, or data bars based on rules

• Red text with a down arrow for performance drops

• Green highlighting for goals met

• Subtle grey for secondary metrics

The key: Don't highlight everything. If everything is important, nothing is.

Shadows and Glow

Create depth and draw attention:

• Add shadows to KPI cards to make them stand out

• Use glow for alerts or critical figures

The rule: One highlight per screen. Maximum.

Here are examples where I use shadows to highlight imagery:

Icons

Universal language that needs no translation:

• 🏠 Home

• 💰 Finance

• 🔍 Filters

• 💡 Tooltips

Icons provide Visual wayfinding. Users scan for familiar symbols, not text.

Here's an example where I use a universal bulb icon for tooltips:

When users hover over it, they get just enough information without feeling overwhelmed:

Text Boxes and DAX

Context without clutter:

• Textboxes are a great way to add affordances to a report. One of the best ways to use them is as tooltips that give extra context without cluttering your design. A simple textbox combined with DAX can do many things.

• In the example below, I'm using two types of text. One simple text provides context about the data, like data source and refresh time. This allows them to assess the information's relevance before diving deeper.

  
  

The text box serves as a Visual cue for users to assess information relevance before diving deeper:

Containers, Padding, and Layout

• Create intuitive visual groupings using the Gestalt principles of similarity and proximity. In Power BI, achieve this by grouping visuals and adding backgrounds to show relationships.

• Add borders around groups to clearly indicate related elements

• Use consistent spacing to either separate distinct content or show relationships between elements

• Maintain consistent alignment throughout the dashboard

• Add borders and separators strategically only where they enhance understanding

  
  

Gestalt principles in action:

• Similarity: Group related elements with consistent styling

• Proximity: Put related items close together

• Continuity: Use alignment to create flow

Here I'm using grouping with background colour to show related metrics:

What This Means in Practice

• You're designing Affordances: making it clear what's clickable, filterable, drillable.

• You're using Signifiers: colour, shape, shadows, icons all helping users understand importance.

• You're Removing friction. A dashboard isn't done when it looks good. It's done when it's clear.

• But don't go overboard. Too many affordances is like too many chefs in the kitchen.

What to Avoid

Overusing bold colours: If everything screams, nothing stands out.

Inconsistent cues: Don't change what colours or icons mean halfway through.

Cramped visuals: White space is a cue too.

Decorating without purpose: Every element should have a job.

In summary

Next time you build a dashboard, don't just ask: "Does this look good?"

Ask:

• Have I shown users where to look?

• Have I made actions obvious?

• Have I indicated what matters most?

And remember

• Visual cues guide attention, behaviour, and understanding without users thinking about it

• Power BI gives you the tools (conditional formatting, shadows, icons, containers)

• Apply Don Norman's principles: make actions obvious through affordances and signifiers

• Design with intent: every shadow, colour, and space should serve a purpose

• Less is more: one highlight per screen, consistent cues throughout

Whether you're designing for executives or analysts, good BI design doesn't just show data. It directs it.

Until next time,

Prathy:)

P. S. I utilise AI to correct grammar and rephrase my sentences.