AI Comes to Analytics

One of the core issues with any analytics application is that it requires a significant amount of knowledge and insight to surface actionable intelligence. There is a small army of expensive analysts that business executives rely on to make sense of all the data an organization collects. However, with the rise of generative artificial intelligence (AI) the level of cognitive ability required to make sense of all the data is about to significantly drop.

For example, Salesforce is adding a generative capability to its Tableau analytics software that, in addition to automatically creating visualizations, will suggest queries that could be relevant based on the data in the application. Dubbed Tableau GPT, this capability will, via a natural language interface (NLI) or using the visualization tools provided by Tableau, make it possible for individuals that normally wouldn’t use this type of application to instantly create all kinds of charts and graphs that might make what was once hard to discern immediately obvious.

Tableau GPT is layered on top of VizQL, the proprietary visual query language Tableau created. It learns what kinds of questions people are asking it to generate VizQL queries than can be viewed as visualizations or fully formed narratives using the Data Stories module the company provides to turn query results into narratives written in natural language.

As advanced as those capabilities may seem, they are naturally only going to be as reliable as the data collected via the application, but if analytics applications become more accessible, then the dominance of the spreadsheet as a tool for analyzing business data may finally be at a close.

Most people who use a spreadsheet to crunch, for example, financial and marketing data do so out of habit. Analytics applications that make it possible to visualize trends have been superior tools for years now. The challenge has been that SQL-based tools used to query data in analytics applications took some time to learn. In contrast, a spreadsheet is readily available to anyone that has licensed Microsoft Office or via their browser access Google Docs.

The trouble is all the data stored in cells within a spreadsheet is prone to error and, for the average user, is obtuse. The more complex the spreadsheet becomes, the more difficult it becomes to manage and, ultimately, surface any meaningful business intelligence.

Salesforce and its Tableau subsidiary have been making the case for analytics applications as an alternative to spreadsheets for years now with mixed success.

Previously, Salesforce rolled out its own instance of a generative AI platform that it has incorporated into Einstein, an AI engine that the company employs across its entire application portfolio. The fact that Tableau is going to employ those same capabilities isn’t a surprise, but the impact generative AI will have on analytics is likely to be profound. Analytics tools that can easily visualize patterns and trends will make it much simpler for the average individual to make fact-based decisions, notes Francois Arjensat, chief product officer for Tableau. “It’s how you compete,” he says.

It’s still early days as far as generative AI is concerned, but organizations that attempt to compete without it will be at a disadvantage they won’t be able to easily overcome.