Dashboard creation process (Best Practices)

This post is a brief summary of the questions you need to ask to build an effective dashboard regardless of the type of dashboard it is.

There are two domains that a dashboard can be used for:

• Analysis (you don't know what you are looking for and that's ok)
• Presentation ( presenting something that should be familiar and highlighting obscurities)

In this blog we will focus on the Presentation domain

So you've been asked by management to build a sales dashboard.  Where do you start?

Let me just first state that a dashboard is a way of viewing multiple pieces of information in the one area/screen.  An effective dashboard is a dashboard where every piece of information has a purpose.  That means if your dashboard contains visualisations then they must also  have a purpose to be effective!

I have lost count the number of times I have been asked to make a "nice" visualisation.  Of course the idea of "nice" changes from person to person, so what I think is nice may be very different to what management regards as "nice". What will capture everyone is "Effectiveness".  A visualisation that is effective will highlight areas of interest and allow you to ask and answer more detailed questions.

Imagine yourself as an architect! 

1. What would an architect consider when asked to design a building?

• What is the dashboard's purpose?
• Who is the visualisations user?
• What is it for? Is there a specific use?
• What is the overall target/goal?
• Is the dashboard to be used as a Navigation tool,  KPI tracker, Summary information, tracking invoice flow?
• How will the dashboard be consumed? Projector, mobile, does it need to be interactive

All have different purposes and will be the driving force for the dashboard design

2. What materials are available and are they reliable?

• Is the data, that the dashboard's purpose relies on, available or could it become available?
• If so, is the data accurate?
• Which data do we need to be able filter?
• Which data do we not need to be able filter?
• Filtering shows relevant and drill-able detail

3. Design & Structure

• Always start your design at a high level and make it easy and intuitive to drill to detail which allows investigation to why something has occurred
• Based on what we need to compare, what type of chart structure would best reveal the data and relationship informed from the purpose and content needed?
• When asked about a cool new graph or if you see a nice graph that you want to use: first of all you need think about what the graph is trying to portray and think about how this could help you look at your data in new/better way.  Always try to use defaults unless you have something better.

And last but not least.

4. Style and Formatting

• Colours, labels are very important as they bring context to the charts and dashboard
• Effective use of colours (You want to be able quickly draw the users attention to recognise important and obscure variances)

To sum up:

• Keep the dashboard purpose in mind at all times during development
• If something is not going to be used then do not waste time and effort putting it into the report.

I hope these tips help you in your dashboard making process and in turn help your company make better decisions.

 James Bingham is a Business Analytics Consultant @ ProStrategy.  Please Like, Share & Comments are welcome! #ProStrategyAnalytics