Learning
This project taught me how to balance exploration with storytelling. I learned how to build a narrative from simple to complex, guiding the audience step by step. I also deepened my skills with event-driven datasets, where frequency, not continuity, is the key dimension. Working with Bitsight pushed me to connect analysis with business communication needs—not just asking “what does the data say?” but also “how do we present this so it resonates with stakeholders?” Finally, building barcode charts (including small multiples and interactivity) expanded my technical toolkit in D3.
Impact
The visualisation became a communication piece for Bitsight, giving them a powerful way to present breach trends to clients and position themselves as experts in cybersecurity risk. By combining analytical rigor with storytelling design, the project helped transform a raw dataset into a compelling narrative.
Challenge
The first challenge was data scope and filtering. Deciding to limit the dataset to US incidents, and which years to include, was not trivial, as every choice affects the story. Another challenge was finding the right visualisation type given the lack of continuous numerical variables. The barcode chart turned out to be the right solution, but it took iterations to arrive there. A third challenge was ensuring the project worked as a client-facing storytelling tool, not just an analysis, which meant extra focus on clarity, interactivity, and usability.
Description
This project was commissioned by Bitsight, a leading international cybersecurity rating agency. They provided me with a dataset of data breaches between 2015 and 2022 and asked me to create an interactive storytelling visualisation. The brief was intentionally open: I had full freedom to decide how to filter, analyse, and visualize the data. That freedom was exciting but also challenging: I had to take responsibility for all analytical choices, from defining scope (which timeframe, which countries to include) to how to treat outliers. Early in my exploratory analysis I realized that while the dataset included global breaches, only the US data was robust enough to use. After validating this direction with Bitsight, I filtered the dataset accordingly. The breaches were event-driven: each incident either happened or did not, with an associated severity score. This reminded me of my earlier work on natural disasters. Since I didn’t have many strong numerical variables, I had to design visualisations that emphasised frequency and categorisation rather than continuous trends. The final narrative unfolded step by step:
1. A line chart showing overall breaches by year.
2. A matrix chart comparing sectors against breach types.
3. A barcode chart visualizing breaches as discrete events across time.
4. An interactive barcode small-multiple view, where users could filter by sector or type and see severity encoded by color. This progression created a story arc, starting simple and culminating in a richly interactive view where users could explore the data themselves.
Topics
Cybersecurity, data breaches · Barcode chart · Small multiples · Interactive storytelling · Severity encoding
Tools
HTML, CSS, JS, d3.js, arquero.js, plot.js
Year
2023
Clients
Bitsight
