Is Your Data Being Logged?

Business Intelligence (BI) work is a continuous cycle of repeating the same tasks back and forth between Excel and SQL. When you find bad data, you have to recalculate where to start fixing it, and this repetition itself becomes a cost.

In environments where data is scattered, like a startup, periodically organizing and extracting the data you need becomes an increasingly heavy burden. So the key is to reduce time cost and choose a tool that can keep up even as the scope of data the company handles expands in the future.

Apache Zeppelin was the first candidate that came to mind, but since it's notebook-based, it was judged to have a high barrier to entry for non-developers—even though it's familiar to developers—and weak visualization. Spotfire, Amazon Quicksight, and Google Data Studio were also examined, but each fell short in terms of references, fit with requirements, or functional limitations.

After the comparison, the strongest option that remained was Tableau. It supports a wide variety of data sources, has powerful visualization, and has abundant tutorials and domestic (Korean) training and references, making it judged as good for reducing dependence on developers while letting analysts and business staff use it together. However, price remained a drawback, and the last candidate, plotly, was concluded to be developer-friendly but lacking compared to Tableau in terms of connectors.

In the end, the criteria were clear.

• Should reduce repetitive work
• Should be flexible for a changing data environment
• Should be usable proactively by non-developers as well
• Should have sufficient domestic (Korean) references and training

Weighing these conditions together, the author chooses Tableau as the tool best suited to the company.