It’s really surprising how BI/analytics tools lack a first-class understanding of the business domain and their needs: metrics, relationships, key analytical workflows, and much needed automation.
These tools center the data producer and despite impressive dashboard designs, leave the manual, tedious consumption work to the end user. Here's a look at how current tools hold back organizations, and why we need the next generation to address these fundamental issues.
1) Data Generation: Repetitive SQL vs Metrics in Code
The first issue is how data is generated. Analysts and data engineers write SQL often repeating the same logic in different places for similar calculations. This results in inefficiencies and introduces risks in data accuracy.
The next generation of tools must shift to metrics in code and APIs serving metrics. This would make metrics consistent, reliable and scalable across the organization.
2) Units of Operation: Datasets vs Metrics and Relationships
A significant limitation of current tools lies in their fundamental unit of operation—the low-level dataset. While datasets offer flexibility to the data producer, they often lead to fragmented, bloated, and inconsistent dashboards as business needs scale.
The next generation of tools must shift focus to entities, metrics, segments, and metric trees as first-class concepts. These tools would allow organizations to directly model and operate on business metrics in a unified, interconnected way, making it easier to align datasets with actual business processes and goals.
3) Analytical Workflows: From Manual to Algorithmic
The most tedious aspect of current BI tools is the expectation on the user to do the heavy lift. Even a simple root cause analysis requires hopping across dashboards, applying filters, performing re-aggregations, and often exporting data to a spreadsheet for additional calculations. This is time-consuming and limits the depth of analysis possible.
Instead, the focus should shift toward automated algorithms and tools that allow users to navigate in-situ—enabling them to run analyses and retrieve insights directly within the interface.
4) Templates and Automation: Unlocking the Power of Standardization
Finally, the current crop of BI tools lack a meaningful approach to business model templates, saved analysis and automation. Without automation, analysts and business users have to reinvent the wheel every time they want to analyze similar data, which results in missed opportunities for timely insights.
The next generation must address this by offering business model codification, saved analysis, and the use of AI to enable “always-on” insights.
These capabilities would enable organizations to operate on metrics and their relationships, streamline analytical workflows, automate repetitive tasks, and deliver insights faster and more efficiently.
