Tableau releases first integration with Einstein Analytics

AI and machine learning capabilities developed by Salesforce are now available in Tableau, marking the first integration between the vendors since the CRM giant acquired Tableau.

Einstein Discovery is now part of the Tableau analytics platform.

Nearly two years after CRM giant Salesforce acquired Tableau for $15.7 billion in June 2019, the two finally released their first integration.

Tableau rolled out its 2021.1 platform update on Tuesday, and included is Einstein Discovery, a no-code tool within Salesforce's Einstein Analytics platform that uses augmented intelligence and machine learning to enable predictive modeling and prescriptive recommendations.

Tableau president and CEO Adam Selipsky and chief product officer Francois Ajenstat first laid out plans for the integration during the vendor's virtual user conference in November 2020, saying at the time it would be generally available during the first quarter of 2021 but not giving any specifics beyond that vague timeline.

With Einstein Discovery now included in Tableau and generally available, users will be able to develop predictive models, create what-if scenarios and forecast within Tableau's environment and without having to write code. The inclusion of Einstein Analytics in Tableau, meanwhile, marks Tableau's introduction of business science, the vendor's term for providing business users with data science capabilities using AI and machine learning.

"This is a big one," Ajenstat said in an interview. "We're trying to create a new market, trying to democratize analytics."

Tableau previously was one of the innovators that ushered in the era of self-service analytics, helping make business intelligence accessible to more than just highly trained data analysts. Now, with the concept of business science enabled by the combination of Einstein Discovery and Tableau, the vendor aims to do the same with complex data science.

"The people who can do machine learning, who can do AI, are few -- it's for specialists," Ajenstat said. "Business science is about lowering the barrier to advanced data science by enabling regular business people to build predictive models, build simulations and do scenario planning. It's all about bringing the power of data science to those business users."

An organization's shipping data is displayed in a sample Einstein Discovery dashboard in Tableau.
A sample Einstein Discovery dashboard in Tableau predicts the likelihood an organization's shipments will be made on time.

Analysts, meanwhile, similarly said that the addition of Einstein Discovery to Tableau is a compelling development.

"This move continues Tableau's five-year-old-plus pursuit of augmented analytics, [and] the addition of Discovery is a solid step forward in delivering ... predictive capabilities that in the past were accessible only to data scientists," said Doug Henschen, principal analyst at Constellation Research.

Meanwhile, the availability of Einstein Discovery in Tableau makes it easy to access and use, and the capabilities it adds are significant for users, according to Dave Menninger, research director of data and analytics research at Ventana Research.

"The beauty is the ease and simplicity with which it is made available," he said. "You don't have to be a data scientist; you don't have to do your job differently. It's just there. Einstein Discovery can unearth insights that an individual might not have thought of otherwise, [and enables] people to focus on determining the correct course of action based on those insights."

Business science is about lowering the barrier to advanced data science by enabling regular business people to build predictive models, build simulations and do scenario planning. It's all about bringing the power of data science to those business users.
Francois AjenstatChief product officer, Tableau

Including Einstein Discovery in Tableau also adds advanced analytics capabilities Tableau was lacking that some competitors already have as part of their platforms.

"With Einstein Discovery, Tableau is certainly among the leaders within analytics and BI tool vendors," Menninger said. "It also positions them to accelerate the use of Einstein throughout more of the platform. However, this does not make them a pure play AI/ML platform vendor. Nor do I think that's where they are trying to go with these capabilities."

Likewise, Henschen said the release of Tableau 2021.1 with Einstein Discovery cements Tableau's position as one of the more advanced BI platforms for augmented analytics. It does not, however, put it on par with vendors such as Alteryx, DataRobot and SAS that deliver deep data science capabilities.

While Einstein Discovery's inclusion in Tableau 2021.1 is the most important feature in the update, it's not the only one. Among the more than 20 new capabilities are:

  • improved connectors to Microsoft Azure that will enable users to access data in Azure SQL Database and Azure Data Lake Storage Gen2;
  • support for Azure Active Dictionary through existing connectors to Azure Synapse and Azure Databricks;
  • a new Extension Gallery that helps users search for connectors and dashboard extensions within Tableau so they can gain access to additional databases and applications;
  • an update to Ask Data, Tableau's natural language query feature, that adds a help center to centralize onboarding and administrative information;
  • a redesigned notification experience that unifies users' shares, comments, extracts and prep flows in a single dedicated space of their choice such as directly in Tableau or in their email; and
  • the ability to create Level of Detail expressions more easily by using context menus or dragging and dropping measures.

"As always, they've packed in a lot of authoring/web authoring improvements, new Ask Data improvements, deeper geospatial analysis and ease-of-licensing/deployment improvements," Henschen said. "On their own, each upgrade could be described as incremental, but the upgrades were plentiful across the platform."

Similarly, Menninger said that even beyond the addition of Einstein Discovery, the Tableau update adds useful new capabilities.

"[Einstein] is obviously the big news, but it continues to surprise me how Tableau hones its user interface and user experience to match users' requirements, for instance with the new notification capabilities and Level of Detail expressions," he said.

While Einstein Discovery and Tableau are finally being integrated, more integrations between Tableau and Salesforce are in the works, according to Ajenstat.

Henschen said that model-building and simulation-building capabilities already on the roadmap will be useful, but that customers might also look toward a data science platform or an enterprise application with machine learning features to add predictive capabilities.

Menninger said he'd like to see Einstein Analytics capabilities integrated into Tableau Prep, the vendor's data preparation tool, and for Salesforce to help Tableau become a cloud-first offering with all capabilities available via the web and SaaS offerings.

Meanwhile, Tableau and Salesforce will work to hone the integration between Einstein Discovery and Tableau, tweaking it to make it even easier and more seamless, as well as add more features, according to Ajenstat.

"This is the first iteration of the integration," he said. "This is the beginning of an exciting journey for us. This is one of those capabilities customers see as changing what they can do with Tableau and unleashing a new kind of analytic power that was not available in the product."

Next Steps

Tableau analytics capabilities aid World Food Programme

Dig Deeper on Business intelligence technology

Data Management
Content Management