Today Tableau has announced the release of the latest version of its analytics software Tableau 2019.1. This includes several new features but also, a new user interface platform called Ask Data. Ask Data is the next generation of self-service analytics designed with the business user in mind. Tableau hopes that it will make analytics accessible to a wider audience.
Enterprise Times had the opportunity to speak to Francois Ajenstat, Chief Product Officer at Tableau Software and see the new product. Ajenstat explained the new feature by saying: “What we have done with Ask Data is integrated it as part of the Tableau Analytics platform. It is automatically available on any data set that is available in Tableau, both live or extracted and in memory. There is zero set up required. The semantics are automatically inferred by Tableau but can be enriched by the users. It leverages Tableau visual best practices and ease of use. It has always been our strength to make the analysis of data more conversational in nature than has been possible before.”
What Ask Data delivers is a step forward in delivering the promise of easy to use analytics for the enterprise and business user. However, one must be cautious and still say that this is a work in progress. For example while the data sets, it interrogates, can be in any language the conversational interface and the taxonomy only supports English in this release. Ajenstat confirmed that other languages are on the roadmap though. Another future update will be a voice interface to the platform.
A modern, easy to use interface
Currently the interface uses an enhanced text interface engine. It allows users to enter queries on their data sets. The clever aspect of Ask Data is that it already understands much of the data that it has indexed. This means that a user can ask a question such as “How many chairs have I sold in France this month?” Ask Data understands that chairs are a type of product and France is a location. It will graphically display the information in a form that is appropriate for the data set using one of the wide variety of charts available. It would initially show this as a bar chart. However, if the user then extended the query by asking for the changes over the last three months, it would change to a line graph as that is better at showing trend data.
As users type in the query, Ask Data provides auto complete suggestions based on other queries completed in the organisation. Once the data is displayed the user is also able to change the graph used and display it in an alternate format. There is a mix of automation and interaction that appears to make it a very simple interface to use. Ask Data also gives the option of changing words such as month with a simple click, provide a choice of month, quarter or annually, for example. This saves the user from always having to type, reducing the number clocks required to see the results of a query.
Other functionality that did not make this release was the ability to do comparisons between different years. Ajenstat confirmed that these would be in a future, possibly next, release. Ask data is able to compare the data between two measures though, for example sales and profit by customer.
Building the knowledge
The Ask Data knowledge base is built up at the load time, this makes that data ingestion a critical point for customers. That semantic layer Ajenstat mentions above contains other elements that increases the intelligent responses that users will get when they ask queries. Ask Data appears impressive. However, only when the users start to test it in earnest will Tableau find out whether their first generation is good enough to meet the expectation.
Ask Data comes with no additional cost as part of Tableau Server and Tableau online. There is a point where the end business user will want to bring in new data sets. This is possible, though probably not best done by the end user themselves.
Sensibly Tableau has also limited some of the functionality for potential performance reasons. While the system can understand common attributes within categories or data, it is currently limited to 10,000 items per dimension according to Ajenstat. This is a temporary limitation and the inference is that Tableau will review it once the platform is live and the impact on the compute power is known.
Using the knowledge
The solution does not end its functionality there. Once a graph has been created, a user is able to drill down into the data. This allows them to discover what lies behind an anomalous data point. Users are also able to share graphs they have created and, usefully, set up alerts if thresholds are breached. This evolves Ask Data from a pretty dashboard to something that can help drive exception reporting for users and of real business use.
It does not yet predict or answers the why question for users though. In fact, Ajenstat was candid by saying that AI often brings back “rubbish” when trying to answer those types of questions. The combination of the drill down capability and the alerting allows the business users to make intelligent choices promptly. Consequently business problems are addressed as they arrive rather than at the end of the month once a report has been analysed.
Experienced users will welcome the ability to move seamlessly from Ask Data into Tableau desktop. This means that while it is a feature that brings analytics to the business users, it is also designed as a simpler to use interface for analytics users. Will enterprises want to pay for a license for Tableau online for more users? Or rather, will customers find that more of their licenses are used with this feature? If the latter is true then this will still have been worth the investment by Tableau. Other companies are looking to simplify their analytics platforms for the end users. PegaSystems recent acquisition of Infruid is an example.
The critically important element for Ask Data to work is properly prepared and clean data. This is one of the areas that Tableau has updated with its new release.
What else is in Tableau 19.1
Tableau highlights four key improvements with the new release:
Tableau Prep Conductor: Tableau Prep, launched a year ago is now used by more than 10,000 organisations according to Ajenstat. It has now launched Tableau Prep Conductor. Ajenstat explained: “The missing link is, once people did that they said :‘ how do I clean my data on a scheduled basis? How do I operationalize my flows? This is where the Tableau data prep conductor comes in to help schedule manage and orchestrate your flows.”
The system allows a user to create schedules to import data sets on a regular basis. Once set up the system will generate reports that can then be reviewed. This helps to automate the process for organisations with that requirement.
Google Ads Connector: Enables Tableau to ingest Google Ads data into the Tableau platform.
PowerPoint Export: Improves the integration with PowerPoint, visualisations exported to PowerPoint are resized appropriately for the slides.
New Mobile Apps: This is a new native app for both Android and Apple iOS. The apps have the same capabilities across both but uses the local design features. They are available later in the month. It also layers the Tableau security into the apps automatically.
Enterprise Times: What does this mean
Ask Data is a very interesting development by Tableau. It addresses two areas of concern that Tableau may have within its customers and prospect base. For existing customers it means that the accessibility and usability of the Tableau platform is far easier than before while still delivering the power of the Tableau analytics platform. For prospects it modernises the interface an delivers a lower cost of adoption for users.
It is also clearly the first step on the future UI for Tableau. It was easy to see how Tableau can build in future functionality that may include AI driven insights to users. Tableau are rightly cautious in approaching this and Ajenstat was realistic in his expectations of the platform.
Within the rest of the release updates the Tableau Prep Conductor is another significant upgrade that also addresses a pain point for many customers. There will also be a whole slew of feature updates and fixes as well besides the ones mentioned above.
If Tableau can increase the usage of its system, it will accomplish three important things. Customers will see a better return on their investment and the customer satisfaction and retention figures for Tableau should increase.