September 2022 Summaries

In the evolving landscape of digital tools, Cube and Apache Superset emerge as compelling open-source alternatives to Looker for data visualization and business intelligence needs. While Cube serves as a headless business intelligence platform focusing on data processing and modeling with support for various data sources, Superset specializes in data visualization and exploration, allowing users to create datasets for richer analysis. Both tools are free under the Apache License 2.0, actively developed, and offer flexibility by avoiding vendor lock-in, contrasting with Looker's more proprietary approach. Cube's data modeling is accessible and adaptable, supporting JavaScript and soon YAML, and its API layer ensures compatibility with multiple data consumers, including BI tools like Superset. This synergy allows Cube and Superset to provide a robust semantic layer for data transformation and visualization, appealing to organizations seeking scalable, flexible, and cost-effective solutions compared to traditional BI tools. Upcoming demonstrations and community engagement highlight the growing interest in these tools as viable alternatives for modern data stacks.

Apache Superset and Preset offer a wide array of visualization types, enhanced by a no-code chart builder that allows users to create visually appealing and functional charts using their team's datasets. The recent update includes comprehensive documentation for the 20 most requested and utilized chart types, such as bar charts, pie charts, and various time-series charts, providing examples, settings explanations, data structuring guidance, and customization options. The documentation particularly highlights the graph chart, illustrating how dataset rows map to visualizations and how data shape and exploration choices influence the generated SQL query. Future plans involve expanding walkthroughs to cover all chart types, with an invitation to explore these visualizations via Preset Cloud and stay updated through their blog.

Preset's blog post explores the application of modern data stack tools like Fivetran and dbt within a BigQuery data warehouse to enhance self-service revenue analytics derived from Recurly data. Fivetran facilitates seamless data ingestion from Recurly into BigQuery, where dbt is employed to transform raw data into specific datasets crucial for visualizing Preset's revenue metrics. The process involves multiple layers of data transformation: initially, raw data tables are refined to enhance clarity by renaming fields and tracking effective dates, followed by creating daily snapshots for comprehensive time-series analysis. The final layer aggregates these datasets into a comprehensive account_subscription_history table, enabling the calculation of key metrics such as Monthly Recurring Revenue (MRR) and Annual Recurring Revenue (ARR). Preset highlights its dataset-centric approach to visualization, providing insights into how these metrics are tracked over time and offering SQL transformation templates to aid in revenue tracking. The blog underscores the utility of these tools in producing detailed visualizations and metrics to monitor revenue growth, demonstrating their practical application in real-world data analytics.