Using interactive data visualization as an effective tool in strategic decision making

This little article describe some highlighted lessons from 2018 regarding to the graphic support to Management Information. The experience comes from The Strategic Intelligence Unit at The Vice-Chancellor's Office, University of Southern Denmark (SDU). It is the collection of the points from the presentations we have held in various contexts during 2018.

Authors: Senior Strategic Analyst Bruno Mølgaard Geertsen, Senior Strategic Analyst Gitte Rønholt Hvidt and Head of Analysis Jacob Jensen

Introduction

In addition to the day-to-day management, there is limited time for analysis when the Board of Directors and other executives must take major strategic decisions in the university's business areas. It is therefore important that data used to support these decisions are easily accessible and quickly understandable.

At The Strategic Intelligence Unit we do this by putting a lot of effort into to making the right interactive visualization for the data we chose for our management information. Our approach to this can be summarized in three main principles:

• Use the right metrics!
• Design matters - beauty is truth!
• Put metrics into perspective - compared to what?

1. Use the right metrics!

Metrics involving strategy, policy and money are important for executive management.

In general: Impact metrics are strong metrics when answering ‘why’ questions in strategic decisions. To help focus on the impact metrics keep the classic 'value chain' in mind: 

Input > Activities > Output > Outcome > Impact

Other issues to be considered:

• Accept a proxy - choose metrics based on information needs and not on data availability
• Use real time metrics whenever possible – data agility gives an analytic advantage
• Think metrics on aggregated levels (e.g. all universities in Denmark instead of just SDU)
• Think metrics on societal key issues that could influence the strategy of the university

2. Design matters - beauty is truth!

Structure and simplify

• Structure information in the dashboard either horizontally or vertically
• Place key metrics at the top, position-charts in the middle, and trend charts at the bottom.
• Keep it simple, i.e. no logos, no fancy functions and no 3D graphics. Less is more!

Consider the use of colors carefully

• Use dark/high intensity colors to highlight specific data - use soft/low intensity colors for the rest.

Only use a few well-known charts (makes the interpretation easier)

• Pre-sorted bar charts are good at displaying benchmark / ranking.
• Line charts are good at displaying development over a period of time.
• Scatter plots are good at relation analysis.
• Maps are an extremely effective tool in geographical analysis.

3. Put metrics into perspective - compared to what?

General issues:

• An interactive design highly enhances analytical usability of the dashboard.
• Whenever two charts / metrics should be compared, always place them next to each other.

Benchmark:

• Comparison between different types of metrics gives a new and a broader perspective
• Comparing organizational key metrics with other similar organizations gives a competitive edge

Time series data:

• Comparing metrics with last year/month is often effective
• Longer time series makes it easy to compare trends
• Points of interest are not limited to high/low values – e.g. crossing curves are equally important.

A few examples on data visualization

Example 1: Never leave a metric to stand alone - compare your key metrics 

Example 2: You only need to use a few classic well known charts (three examples)

Example 3: Dashboard Design - showing key indicators on Student Activity

Example 4: Dashboad Design using scatterplots showing patterns for Student Activity

Example 5: Dashboad Design using maps showing mobility

Our conclusion

Our experiences so far: Clear interactive data visualizations provide top-management with an effective analytic tool, allowing for quicker and more extensive analyzes. Focus on their needs and give them quickly understandable trends and key figures to help them take strategic, evidence-based decisions.

Your experiences?

Please write your comments - or contact us for further sharing of knowledge on data visualizations :-)