
Designing dashboards stakeholders actually use is one of the biggest challenges in analytics. We can build the most technically impressive report (perfect blends, advanced formulas, gorgeous charts) and still hear: “I don’t understand this.” It happens more often than most people admit.
The gap between a dashboard that analysts love and a dashboard that stakeholders use is rarely a technical one. It’s a design and communication problem. The good news is that Looker Studio gives us the tools to fix it. We just need to approach dashboards with a slightly different mindset.
In this post, I’ll walk you through design principles that I use with real client projects, using the Spotify 2023 and Superstore Products datasets to illustrate them. My goal is to help you build dashboards that stakeholders actually open, explore and use to make decisions.
1. Begin with a question, not a chart
A dashboard is not a gallery of widgets. It’s a tool to answer questions.
When I start designing, I don’t ask:
“Which chart should I use?”
I ask:
“What decision is this stakeholder trying to make?”
For example, the Superstore Products dataset includes Category, Subcategory, Gross revenue and Profit. Instead of jumping into graphs, ask:
- Which categories generate most revenue?
- Which subcategories are profitable but under-invested?
- Which regions consistently underperform?
Once you have questions, you design for clarity instead of aesthetics.
A good rule: every chart should answer one question.
2. Prioritise fewer metrics with higher context
Stakeholders are not analysts. They don’t want to decode ten metrics. They want to know what matters.
Let’s compare two versions of a simple KPI block using the Spotify dataset.
Less effective:
- Streams
- Track count
- Average BPM
- Energy %
- Danceability %
It looks impressive, but it raises more questions than it answers.
More impactful:
- Total streams
- Top 3 artists
- Growth vs previous period
Stakeholders understand narrative, not numbers. Always design KPIs to tell a story at a glance.
3. Put the most important insight first
Dashboards are scanned, not read. When you open a Looker Studio report, your eyes go to the top-left, then top-centre. That real estate should contain the most valuable information.
For a sales dashboard using Superstore Products, you could design as follows:
- Top-left → Revenue summary (e.g., month-to-date + variation vs last month)
- Top-centre → Profit summary
- Top-right → Key opportunities or warnings
This creates a “decision layer” before any detail appears.
4. Use friendly language, not technical labels
Stakeholders don’t think in terms of database schema. They don’t want to see “Gross_revenue” or “artist_count”.
Use natural labels:
- “Revenue”
- “Streams”
- “Total products sold”
- “Average order margin”
Naming is design.
When in doubt, ask yourself:
“Would a person with zero Looker Studio experience understand this?”
If not, rename it.
5. Choose chart types that match decision-making
Every chart has a job. Use them accordingly.
- Time series – performance over time
- Bar charts – comparisons
- Treemaps – proportions and hierarchies
- Tables – operational lookup
- Scorecards – KPIs
One of the most common mistakes I see is treating dashboards like art. A gradient pie chart might look beautiful, but it won’t help anyone choose which subcategory deserves budget next quarter.
A treemap using Subcategory + Gross revenue in Superstore works because it shows relative importance instantly. You don’t need to explain it.
6. Make dashboards interactive, not static
Interactive elements convert dashboards into tools.
In Looker Studio, there are three simple features stakeholders adore:
- Drill down
- Filters
- Date controls
Example with Spotify:
- Dimension 1: artist(s)_name
- Dimension 2: track_name
Enable drill-down so users can zoom from artist to songs without editing the report.
For Superstore:
- Dimension 1: Category
- Dimension 2: Sub-Category
Stakeholders click → they explore.
They feel ownership instead of frustration.
7. Reduce noise: remove anything that isn’t helping a decision
This is the hardest skill to learn. A dashboard with 12 charts looks impressive, but stakeholders use… maybe two.
Ask this for every element:
“If I delete this, does the decision become harder?”
If the answer is no, delete it. Removing noise isn’t simplification, it’s strategy.
8. Give your dashboard a narrative
Dashboards should read like a story:
- What is happening?
- Why is it happening?
- What should we do next?
A design I often use: Headline → Context → Breakdown → Detail
Example (Spotify):
- Headline: Total streams + % vs last period
- Context: Top artists
- Breakdown: Time series
- Detail: Table of tracks
This structure works because it mirrors how stakeholders think.
9. Use colour intentionally
Colour is not decoration; colour is information.
- Use green for “better than expected”
- Use amber for attention
- Use red sparingly, only when action is needed
- Keep brand colours consistent
A simple palette beats rainbow chaos every time.
10. Close the loop: document the dashboard
Stakeholders forget. Teams turnover. A dashboard without documentation becomes abandoned in weeks.
Add a short section:
- What data sources feed it
- How metrics are calculated
- What actions users should take
This takes minutes and saves months.
Final thoughts – Designing dashboards stakeholders actually use
Designing dashboards stakeholders actually use is less about visual flair and more about empathy. Your job isn’t to display every insight, you guide the viewer toward the few that matter.
Start with questions. Prioritise clarity. Use interaction sparingly but thoughtfully. And whenever possible, bring your user into the process early. They’ll tell you what they care about, and that will make every design decision easier.
The dashboards that get shared and bookmarked are not the fanciest. They are the ones that answer questions confidently.
FAQs – Designing dashboards stakeholders actually use
-
How many KPIs should I show at the top of a dashboard?
Usually three to five. Enough for direction, not enough to overwhelm.
-
What is the best first chart?
A time series or high-level scorecard. Stakeholders want context first.
-
Should I hide complicated tables?
Not hide, push them to the bottom. Tables are for detail, not narrative.
-
Do dashboards need instructions?
Yes. A short description of controls or drill-downs massively increases adoption.
-
How do I know a dashboard is working?
People talk about it, reference it in meetings, and ask for extensions, not explanations.



