How to Design Effective Analytic Dashboards that Look Incredible in 8 Steps
Dashboards are the go-to data hubs that businesses rely on for decision making. They provide visual answers to business questions so people can act fast on opportunities and issues. Dashboards are powerful tools.
But they're not always like that, are they? Dashboards can often end up being quickly-thrown-together summaries of data that someone asks for last-minute. Then there are the conflicting views on what's most important and what should be included that can make dashboards a mess. And it's easy to fill up the white space with data that's related, but not actually needed, so it looks less sparse.
But fear not. Here is a step-by-step guide to designing dashboards that are effective and look great too.
1. Who is this dashboard for?
If you answer 'Marketing', then that's not specific enough - by a long way.
You need to know which people, in which roles, are going to be using this dashboard. You need to know the precise question they are trying to answer.
It's your job to present the data in a concise manner that's immediately understandable. And preferably, your dashboard will look great too. That will keep people engaged.
With a clear question to answer, an understanding of the users' roles, and knowing what data they need on a regular basis, you can tailor the dashboard design to make it effective in answering questions fast so they can make accurate business decisions.
2. Build the right type of dashboard
Did you know there are different types of dashboards?
Here are the three primary types you could design:
Operational Dashboard
Operational dashboards help managers track the day-to-day running of different operations so they can make short-term decisions to optimise performance.
Operational dashboards should host live, near real-time reports on each aspect of each operation. For example, the Marketing Manager might want to oversee the marketing operations, but that means having visibility of the department budget, Google AdWords performance, social media engagement, the number of leads generated, the performance of website landing pages etc. Much of this requires pulling on third-party data using APIs.
Each marketing operation should be put on a separate dashboard tab and the different elements of each operation can be displayed in summary reports. To prompt action, set up alerts for when values fall outside of predefined parameters so the manager knows when to act.
Strategic Dashboard
Build strategic dashboards for executives who need big-picture oversight. This type of dashboard should include high-level summary reports and provide instant information on how the business is performing against its Key Performance Indicators (KPIs).
Strategic dashboards are designed to help executives measure high-level performance and provide a global view of the business functions. It helps them monitor progress toward goals and supports long-term decision making. But you also need to enable drilling so that the executive can explore some of the numbers behind those summary reports in order to understand what is driving those performance results.
You will need to pull from multiple data sets to provide all the information required for the strategic dashboard.
Analytical Dashboard
Analytical dashboards are designed to be fast, highly interactive, and link all your data together so the users can explore the data in more detail. These dashboard types are ideal for people who are comfortable with data analysis. Enable brushing, series selection, filters, and drills so the users can uncover fresh insights to understand the what, why, when, and how of the data.
The analytical dashboard could display either strategic or operational data, but the aim is for the user to be able to dig into the data further rather than simply get an overview of what is happening. Instead of the day-to-day running or a very broad overview, the analytical dashboard allows data-savvy users to make strategic medium-term decisions.
3. Choose the right chart types
Dashboards are a collection of reports on a single page, so choosing the right chart types for your data visualisation is essential. Below is a quick guide to just some of the chart types you can choose from. Use the type that best communicates your data insights.
You can find this diagram and other helpful tips in the Dashboard Best Practices Guide - download it here.
4. Nail the layout
How you organise a dashboard can affect how it is read and how effective it is in answering those business questions. You want to reveal the data story - the what, who, where, and when of the data - to help the user gain insights.
Tell the story in order
Put your reports in an order that makes most sense for the data story. You want to lead people to draw their own conclusion. Position the most important information at the top of the dashboard so the user sees that first. In western cultures, we are taught to read from top to bottom, left to right. So follow that pattern in ordering the data - most important in the top left to least important in the bottom right. In other countries, the dominant reading pattern may be right to left. Use the convention of your audience.
Make space
Give your most important reports the most space. Make those reports bigger than the others (if appropriate) and add some white space around them. Adding some white space around important reports will make the dashboard users pause to study it - the space slows people down a little.
Keep reports legible
Above all, make sure all your reports can be read easily. Don't squash them or squeeze them into awkward spaces just to fill the blanks in the dashboard.
Group your data
Group the reports that are related. You could use background colour, spacing, or alignment to visually group related reports. If there are a large number of reports relating to the same topic, you could create a sub-tab on the dashboard.
Don't make them scroll
Keep all your reports above the fold where possible. Don't make your dashboard users scroll to find the information they want. They want information at a glance. If you have a lot more data than fits on one page, consider breaking out the data into sub-tabs where associated reports can be grouped.
5. Make it crystal clear by adding context
You could create visually stunning reports and design the coolest dashboards, but if the information can't be understood immediately, you've created art, not a helpful tool for business.
Use the simplest visual for representing your data. It might not be the sexiest, but if it communicates the insights faster, use it.
But it's not just over-fancy designs that hinder understanding; lack of information does too. Context is crucial. Make sure you label all of your reports. Include scales on the axes, titles for each report, and legends for colours and patterns. Enable tooltips so that specifics can be revealed where needed. You could even add a short sentence to describe the key findings. Whatever makes the data most digestible, use it. But don't clutter the dashboard. Give the tools that are needed - no more, no less.
Read the blog post How to Enhance Your Data Visualizations with Context for a more detailed look at this.
6. Keep everything consistent
One of the biggest challenges is making sure different people have the same understanding of the data. Avoid confusion by building consistency into your dashboards.
If you use the colour purple to indicate Gross Profit on one report or dashboard, then use purple to represent Net Profit on the next, you will confuse the dashboard user. Keep colours consistent to help the users across the business form associations with data. Do the same for labels, metrics, and calculations. Consistency breeds trust, which nicely brings us to...
7. Is the data trusted?
Is your data trustworthy? If the data on the dashboard doesn't match the data the user has, you'll end up breeding distrust. Work with the users to define the source of truth and work from there. Use only quality data and keep the data clean. This all comes back to robust data governance. Business Intelligence (BI) projects fail fast when the data is poor, and broken trust is hard to repair.
Maintaining quality data is a continuous task, but a vital one. Data is added all the time and the same quality checks need to be done on new data as was done on the initial data used for the dashboard.
Build in a feedback loop that empowers your dashboard users to flag data quality issues. With Yellowfin Tasks, users can take screenshots of reports and dashboards, add annotations to them, and assign the task to you or fellow analysts so you can quickly resolve any data issues.
How often you refresh the data plays into trust too. If the data isn't refreshed often enough, users will search elsewhere for their up-to-date data insights. Refresh the data often, but not more than is needed or you will slow down data retrieval.
Read more about trustworthy data in Intelligent BI: Why you need trustworthy business data.
Having data at your fingertips is a powerful tool, but not everyone wants to log in to the BI platform every day. That's why it's important that you enable users to share their dashboards and data. Data needs to be accessible to be useful.
Have you set up your dashboards to be viewed on mobile devices? Many users want data on the move or in meetings. Make dashboards available where they're needed.
For those who just need the data once a week or once a month, why not set up broadcasts that send the reports via email or within the BI platform as a notification. Users should also be able to embed the insights into web pages and emails. Consider sharing specific data insights in an eye-catching infographic. Engaging formats encourage others to share the data.
The methods for sharing data are numerous, so make sure your dashboards can be distributed.
In conclusion...
Clarity is everything.
Design effective dashboards by:
- Designing the dashboard for a very specific user or user group
- Building the right dashboard type for the job
- Choosing the right chart types for your reports
- Laying out the dashboard in a way that imparts information logically
- Adding context for clarity
- Keeping colours, calculations, labels, and metrics consistent
- Making sure the data is trustworthy - clean, complete, and high quality
- Enabling users to share their dashboards, insights, and data
>> Download the full 10-point guide to designing dashboards here