Interactive visualisations are a powerful tool for data exploration and communication, allowing users to engage with data in a dynamic and intuitive way. They enable users to manipulate and explore data through various interactions, such as zooming, filtering, and drilling down into specific data points.
I decided to create this SvelteKit application to share my approach to building interactive data visualisations that are user-focused, accessible, and effective. Here, I document techniques, design decisions, and lessons learned, while developing visual tools that help users explore and understand complex datasets. All examples are taken from my career history.
Guiding Principles
My approach to creating interactive visualisations is grounded in four guiding principles.
I always begin by asking a series of fundamental questions: Who are the intended users of the data? Who will be the ultimate customer of the visualisation? What decisions are these users trying to make? These questions anchor the entire design process, because data is only meaningful if it drives action. Without clarity on the user and their goals, visualisations risk becoming decorative rather than functional.
Visualisations should be designed to help users reveal hidden structures and relationships within their data. They should make abstract data and complex phenomena perceptible, transforming numbers and codes into forms that can be intuitively grasped. By enabling exploration and interpretation, visualisations support users in gaining deep insights and understanding from their data.
Effective data visualisation design balances elegance and approachability with the ability to explore deeper complexity when needed, serving as a bridge that translates algorithmic knowledge into practical business insights; by making strategic design choices that respect human cognitive limits while capturing the richness of the data, complexity becomes a source of profound understanding rather than confusion.
Visualisation is an iterative, evolving conversation with both data and users, generating new questions rather than providing a single definitive output; this process acknowledges that design solutions rarely succeed on the first attempt and recognises that data is not neutral but embedded within social contexts, power structures, and potential biases, requiring critical questioning of its sources and ethical implications to produce meaning, build trust, and support informed decision-making.
The Three Core Pillars
These three guiding pillars - the Why, What, and How - form a structured approach to designing impactful visualisations. They ensure that every visualisation is grounded in user purpose, driven by data integrity, and executed with clarity, accessibility, and design excellence.
Before writing any code or designing any visuals, I prioritise understanding the purpose, context, and user goals. I ask why the visualisation is needed, who the users are, and when and how they will use it - whether under decision-making pressure in a policy meeting, on a mobile device in the field, or by the public during a crisis. This editorial thinking ensures I choose the right angle, framing, and focus to tell an honest and insightful story.
Next, I immerse myself in both the data context and the user context. This involves rigorous exploratory data analysis to understand data types, distributions, and structures, while engaging stakeholders through interviews, workshops, and usability testing to map their goals, mental models, and constraints. Trustworthiness here is rooted in integrity, ensuring designs are driven by the data itself and deeply aligned with user needs rather than pre-conceived narratives.
Finally, I translate the purpose and analysis into a thoughtful design. I choose visual forms that best support user tasks, apply deliberate design choices rather than defaults, and ensure accessibility through colour-safe palettes, keyboard navigation, and responsive layouts. By incorporating interactivity for insight and maintaining elegance in typography and composition, I create visualisations that are clear, credible, and memorable.
Map design is a critical aspect of data visualisation, involving specific principles and operating within a broader context of general design considerations, from visual elements to interactivity, all aimed at effective communication.
Creating interactive visualisations is a multi-stage process with various design choices. Different chart types significantly influence both data comprehension and user interaction by leveraging various aspects of human visual perception and enabling different levels of data exploration.
Visualisations are essential for allowing users to understand a dataset's key features from different angles and unearth confirmatory or enlightening discoveries. Visualisations are integral to making knowledge claims accepted and credible by communicating exactness and objectivity.