Salesforce introduced a new feature to Einstein Analytics today called conversational queries. It’s not conversational in the Alexa sense of conversation, at least not yet, but it does recognize the most common phrasing as you’re typing, providing an automated way to build queries and access data.
“Now, with Conversational Queries, users can type phrases related to their data — such as “show me top accounts by annual revenue” or “rank accounts decreasing by annual revenue and billing country” — and instantly view answers in automatically configured dynamic charts,” Amruta Moktali, VP of Product for Einstein Analytics wrote in a blog post announcing the new feature.
While the tool has traditionally worked well for technical users to build queries, it requires a broad knowledge of how to find the information you’re looking for and fashion it into an actual query. By putting the queries in plain language, helping you build the query automatically, then building the charts based on the automated phrasing, it puts the ability to access analytics into more hands.
“The way we are doing it, we know the data structure, so rather than [forcing you to do the] guesswork, we actually suggest the correct query, making sure you are not making a mistakes,” Moktali explained. The idea is to reduce the trail and error that typically goes into building a query and make it a more efficient process with fewer clicks and false starts.
“Traditionally, the process of creating charts has taken an average of 12 clicks and required a deeper understanding of chart building — how to create filters and what measures should go on each axis,” Moktali wrote. “Conversational Queries offers a new way to explore data and get answers to questions faster, eliminating clicks and the training required to create and drill down into charts,” she added.
While the analytics tool was renamed Einstein Analytics last year from its original name, Salesforce Wave, the new functionality is not using artificial intelligence to make it work. Instead, Moktali says, it simply relies on the underlying metadata to build the queries using a simple old-school search technique.
This isn’t trying to be fancy in a technology sense. It’s not even taking advantage of natural language processing or any type of artificial intelligence. It’s simply exploiting the metadata every customer has to create as part of their Salesforce install process and building the queries based on that underlying information.
Metadata is defined as data about data or how you identify your data elements. It’s been long used in online help and search tools inside enterprises to make it easier to locate documents in a search context, but this is an interesting way to use the metadata to speed up query building and expose information more quickly.
The new automated query building feature is available in beta today. It should be released by the end of the year, Moktali says.