What data visualization tool should I use?

At some point, every organization needs to present its data and metrics in a structured manner. The process of requesting the IT team to manually extract data, dump it into Excel, and create charts is not scalable. For larger organizations with more data, there is a growing demand for accessibility and handling large data volumes. The process must be automated, run stably, and often involve combining data from different tables or even multiple sources.

As a result, it becomes crucial to choose a solution for this – a reporting tool where you can craft dashboards, visualizations, and set up reports that serve as the "one-source-of-truth." With a plethora of tools available, making a choice is no easy task. Additionally, much of the software offered tends to be relatively costly. Once you commit to a specific tool, switching becomes a cumbersome process. In other words, it's essential to make the right choice from the outset.

Below I will highlight some aspects I think are important in this process.

The first step is to be clear about your expectations and goals. Addressing these questions is crucial for navigating the subsequent steps:

• How do you envision your organization utilizing the tool?

• Will business users engage in a lot of self-service reporting and data exploration, or will the data be statically consumed with pre-defined filter and breakdown selections?

• How is your data structured?

• Do you anticipate the need for extensive aggregations, table joins, or even the integration of data across different warehouses? Alternatively, is it a more limited selection of tables that should be considered?

• How do you ensure that you have personnel capable of maintaining and developing the tool?
  You will need to upkeep and evolve your reports, whether through external consultants or in-house expertise.

Below, I will briefly present some of the biggest players in the world of reporting software.

Power BI

Power BI is Microsoft's BI solution, seamlessly integrated into the Microsoft Office Suite, which is undeniably its main unique selling proposition (USP). It allows advanced data merging and manipulation through its 'tabular models,' offering a plethora of connectors to most data warehouses. However, building dashboards can be somewhat challenging, and the overall look and feel appear somewhat dated. Perhaps most crucially, users need familiarity with the not-very-intuitive DAX language to perform more advanced calculations.

Tableau

This tool stands out as a powerhouse among reporting tools. It boasts remarkable flexibility in creating and designing visually appealing dashboards. It is exceptionally responsive, capable of handling vast amounts of data, and facilitates the easy creation of various custom metrics and calculations. However, developing with the web interface is not the most user-friendly experience. Additionally, it comes with a relatively high price tag, making it less suitable for smaller organizations that might not utilize all the functionalities the tool offers.

Looker

Looker, owned by Google, stands apart from the previous giants in certain aspects. It places a strong emphasis on data consistency, achieved through the creation of a central data model using their 'LookML' language. While it allows for the creation of appealing dashboards, its standout feature is its self-service capabilities. Looker is most beneficial for mature organizations with a high level of data literacy; smaller organizations or those requiring only a few dashboards may find the tool too complex. The 100% web-based platform makes it exceptionally lightweight, and the built-in versioning is an additional perk.

Looker Studio (former Google Data Studio)

Looker Studio, a free tool from Google, despite its name, differs significantly from the 'real' Looker. It is tightly integrated into the Google Suite, particularly with Google Analytics, making it convenient for visualizing this specific data. However, I have rarely encountered any clients using Looker Studio for anything beyond ad-hoc data explorations.

This aligns with its intended purpose. While it is free and offers attractive charts with excellent design features – notably more user-friendly than Power BI – its capabilities are limited when it comes to more advanced tasks like creating metrics or joining datasets.

Qlik, Spotfire, SAS Visual Analytics, SAP Crystal Reports, ...

In addition to the previously mentioned tools, there are numerous others in the market. While each has its advantages, they tend to have a relatively low market share. These tools are backed by large corporations, ensuring reliable solutions that are often well-integrated with some data warehouses. However, they lack versatility in terms of connectors, and, to be honest, their visuals are mostly quite primitive.

Open source

Finally, there is also the option to code your own apps using Python, R, or Django. There are various alternatives and packages available; personally, I find the R-Shiny package to be quite appealing. This approach can be considered if you have only a few metrics and charts, if things are relatively stable, and changes are infrequent. Of course, you'll need someone with coding skills, but the same holds true for other tools – you need someone who can navigate and utilize them. Don't be concerned that this might lock you into something; these tools are well-known, and there are plenty of developers available who can maintain, troubleshoot, and further develop them.