Where exactly does data visualization end and analysis begin?

The line between Analytics and Visualization software offerings continues to blur. Data visualization software can form part of already sizable (and increasingly unwieldy) tech stacks and workflows. In response, visualization tools are slowly being integrated within broader, diverse analytics software product suites designed to support data management and storage.

Visualization depends upon reliable data, which often requires pre-processing. For that reason and others that relate to data and the analytics performed on them, it has become increasingly difficult to call a data visualization tool just that. Freestanding software tools do not create visualizations in isolation, and it does not make sense to separate visualizations from data management or broader analytics efforts.   Visualization ultimately relies on data management, data transformations, and ancillary analysis. Creating good visualizations presupposes an understanding of the data, an ability to ask reasonable questions of it, and an appreciation for the motivation of those questions. Anyone involved with data visualization needs to understand the basis of inquiry that led to the representation of the data in the first place – and have a view on the results.

All of which should be supported by software.

I’ve been involved in several dashboarding projects of late. In the process I discovered that advances in data visualization fostered by innovations in technology provide greater flexibility in developing and publishing visuals. Open source software (such as that made available by RStudio) supports customization, with CSS and Java script for enhanced graphic design. The resulting dashboards can be hosted on dedicated servers in the Cloud and wrapped with security features.   Many commercial software alternatives often better handle challenges brought about by Big Data, allowing visualizations to update in real time and incorporate data refreshes.

The market for BI Software and tooling is crowded, with many offerings that do different things well. Open source alternatives, whether developed in Python, R, D3, Shiny and Dash, all have solid graphics capabilities, and now vie with commercial alternatives. It’s getting easier to picture a world of consolidation, in which the larger platforms incorporate visualization tools more seamlessly. Witness the partnerships that have been cropping up, including those between Trifacta and Tableau, Alteryx and Power BI, and Alpine Data and Tibco leading the way.

Many business end users are content with Excel for charting. Yet Excel remains clunky and impractical for visualizations in the era of Big Data. Those with greater aspirations search for something else. No matter the software used to develop them, the days of generating charts in Excel then copied or written out to PowerPoint ought to be a thing of the past. 

Thanks to advances in software it is a great time to be involved with data visualization for those exploring data – whether in support of business and finance, journalism, urban planning, or any number of other applications. Software is leading the way.