- ✓Industry-standard data visualisation platforms such as Tableau and Microsoft Power BI enable analysts to connect to data sources, transform data and create interactive dashboards without requiring deep programming knowledge.
- ✓The choice of chart type is not arbitrary: bar charts, line charts, scatter plots, maps and other visualisation types each communicate different kinds of information and relationships, and using the wrong type can actively mislead your audience.
- ✓Design principles including colour choice, labelling, gridlines and layout have a significant impact on how easily a visualisation can be understood: good data visualisation is as much about design as it is about data.
- ✓Interactive visualisations that allow users to filter, drill down and explore data themselves are increasingly expected by business audiences, as they enable more self-directed and exploratory use of data.
- ✓The ability to produce clear, accurate and compelling visualisations is a highly marketable skill: data storytelling, the practice of presenting analytical findings as a coherent narrative supported by visualisations, is one of the most valued capabilities in modern data roles.
Listen to the full episode inside the course. Enrol to access all 80 episodes, plus assignments, tutor support and Student Finance funding.
Start learning →Alex: Welcome back. Today we're getting hands-on with data visualisation tools, and this is the lesson where theory starts to become practice. Sam, what tools do learners need to be aware of?
Sam: There are quite a few in the market, but the industry has coalesced around a relatively small number of dominant platforms. Microsoft Power BI is very widely used in corporate environments, particularly in organisations that are already heavily invested in the Microsoft ecosystem. Tableau is probably the most feature-rich dedicated visualisation tool and is widely used in analytics-focused organisations. Qlik is another enterprise platform. And for those who prefer a programming approach, Python has excellent libraries including Matplotlib, Seaborn and Plotly, and R has ggplot2, which is genuinely beautiful for statistical visualisation.
Alex: What's the typical workflow when you're creating a visualisation from data?
Sam: It starts with connecting to the data source, which might be a spreadsheet, a database, a cloud data warehouse or any number of other sources. Most tools have built-in connectors for common data sources. Then there's a data preparation step where you clean the data, handle missing values, create calculated fields and shape the data into the structure the tool needs. Then you move into the actual visualisation design: selecting chart types, choosing the fields to plot, configuring axes and legends and colour scales. And finally you work on the presentation layer, adding titles, context, filters and interactivity to make the visualisation usable by its intended audience.
Alex: The data preparation step seems like it might take a lot of time.
Sam: It typically takes the most time. In the real world, data is rarely clean and perfectly structured when you receive it. You'll encounter missing values, duplicates, inconsistent formatting, data that's in the wrong type and relationships that need to be inferred rather than explicitly stated. Learning to prepare data effectively, using tools like Power Query in Power BI or Python's pandas library, is a core skill for anyone working in data.
Alex: Let's talk about the design decisions. How do you choose colours, for example?
Sam: Colour is one of the most powerful tools in data visualisation but also one of the most commonly misused. Sequential colour scales, ranging from light to dark in a single colour, are appropriate for data where values have a natural order from low to high. Diverging scales, with two different hues meeting in the middle, work well when you want to show values above and below a central point like zero. Qualitative palettes use distinct colours with no implied order to distinguish between categories. And critically, you should always design for colour blindness: avoiding red-green combinations and using symbols or patterns alongside colour where possible.
Alex: And what about designing for a specific audience?
Sam: This is where a lot of data visualisations go wrong. The analyst who created the visualisation knows all the context, all the data and all the caveats. The audience typically knows none of those things. A visualisation designed for an analyst audience needs very different treatment from one designed for a general management audience or for public communication. The level of detail, the complexity of the chart types, the amount of explanatory text: all of these should be calibrated to the actual audience.
Alex: Brilliantly explained. Thanks Sam. We'll look at data specialist roles next.