In recent years, data visualization has been a talking point on everyone’s lips! It is the sublimation of the data democratization process, enabling both small and large enterprises to have their business strategies led by data and insights rather than gut feelings or pre-fixed ideas. Data visualization is the discipline which gives life to the vast amount of data that we have at our fingertips. The benefits are numerous! It allows businesses to understand and act on the data they track. It allows journalists and scientists to give rigour and explanations to their findings. Data visualisation provides thousands of people all around the world with fun in manipulating and experimenting with data. It supports communities around the world to tackle worldwide issues such as poverty, health and environment. It elevates data to a new dimension and power like never before!
Nowadays there is an explosion of new roles which are tied specifically to data visualization (data visualization engineer, analyst, specialist, director, etc.) and more and more data visualization enthusiasts are gathering in communities to find inspiration, try new methodologies and have fun (#makeovermonday and #datafam are the best examples).
There are several evolving BI tools available to easily craft stunning visualizations. Recent examples include tools like Power BI, Looker and Tableau, all eager to upgrade and sharpen the claws of their products in such a challenging and demanding market.
This post aims to answer a simple question: what makes a visualization successful and effective? Short answer: not just the chosen tool! Achieving the purpose-audience fit is what makes a visualization great and effective. What is the purpose-audience fit? Taking the notion from the growth vocabulary ("product-market fit"), a purpose-audience fit visualization is a viz which successfully links the main core messages with the audience and final users’ characteristics in terms of demographics, knowledge of the topic, roles in the company/society, time available to digest the viz and personas.
There are four main archetypes which include the most common purpose-audience fits. Each archetype uses a different methodology to achieve that fit.
1) Performance Scorecard
Purpose – Performance Scorecard aims to convey a few key messages to an audience where the span of attention and the time to consume information is limited.
Design – Performance Scorecards are sober, essential and straight to the point visualizations. In a glimpse users know what is happening in the business. This means the design should be minimal, including elementary simplistic graphs with numbers in big format, which are useful to synthesize the message on the dashboard. Attention to colour is less important, but can help to differentiate between positive and negative trends and performance (the classic good-green and bad-red).
Audience – People with strong knowledge of the topic but with limited time to digest the information.
Examples – The Performance Scorecard archetype can include different kinds of dashboards. It could refer to the financial/executive scorecard, a dashboard widely used by top management to skim the health of the business and the major KPIs. It could be a top marketing dashboard, where marketers can immediately spot the channel’s traffic performance, campaigns ROI and customer insights, or it could be a healthcare scorecard with the hospital achievements and performance, etc.
2) Self-Service Tool
Purpose – Self-service tools enable users to actively manipulate and consume the visualization to find their own conclusions and results.
Design – Self-service tools usually guide the final users with clear instructions and clear steps on how to use the visualization. This archetype is embedded with arrays of filters and parameters to modify in order to be tailored to the needs of the user. Colour importance is low, text boxes support having a user-friendly structure with clear guidance on how to customize the analysis and the results interpretation.
Audience – The final users of self-service tools have a great understanding of the topic, they are interested in it and they have time to deep dive on the subject and to use the visualization for their own findings and purposes.
Examples – Self-Service tools are often generated by business analysts to empower the end-users to run their self-analysis in the business side. Numerous examples can also be found on the internet in case you need to fill in a form or use filters to get the results. A famous example (with traits of Performance Scorecard) is the John Hopkins University Coronavirus Resource Center dashboard.
3) Storytelling Visualizations
Purpose – Storytelling visualizations are used to support, reinforce and facilitate the comprehension of insights and messages. They enable consumers to understand and absorb the conclusions of the author.
Design – Storytelling visualizations use a wide range of graphs to convey insights. Usually these graphs are built following the common best practices that data science has established (line chart for timelines, bar charts for dimensions split in different categories, scatterplot for quantitative vs quantitative analysis, etc). Colours are paramount to distinguish the different dimensions/categories and to facilitate the reading of the graph/story flow.
Audience – The audience for this archetype usually has a low to medium understanding of the topic and not enough time to retrieve the information by themselves.
Examples – Storytelling graphs can be found in almost every presentation which entails analysis and explanation of the work done. In recent years, the media industry has successfully exploited the power of well-structured visualizations to improve the quality of their articles and messages. An explicative example of a media actor which is successfully embracing the data visualization opportunity is the New York Times.
4) Data Art
Purpose – Data has enormous power nowadays and can be used as an effective tool to wow the audience. The main objective of this archetype is to both surprise and fascinate the viz consumer and to strongly reinforce the creator message.
Design – In this archetype all common standards are abandoned. There is absolute freedom on how to display data, what colours to use, what images or decorations to apply, and the tone of the visualization. In Data Art archetype all best practices on data visualizations are bent to the creator design style and the message that the viz aims to convey.
Audience – Usually data art visualizations target the general public, who may or may not have knowledge on the subject explored.
Examples – Data art visualizations can be found in journalistic essays and articles, in politicians’ speeches and specialized websites.
Conclusion
It is always challenging to face countless situations that we find in real life and to simply allocate them into rigid categories. Nevertheless, the four archetypes listed in this post cover most of the use cases that a good data viz creator should know during the ideation and building phase. Understanding them is key in achieving a successful purpose-audience fit and therefore an effective visualization.
