Automatically Build Views with Ask Data
Ask Data introduces an entirely new way to interact with your data, letting you type a question and instantly get a response right in Tableau. Answers come in the form of automatic data visualizations, with no need to manually drag-and-drop fields or understand the nuances of your data’s structure.
Ask Data lets you ask sophisticated questions naturally, with support for key analytical concepts like time series and spatial analysis, and an understanding of conversational phrases like “last year”, “earliest”, and “most popular”.
In Tableau Server or Tableau Online, navigate to a data source, and then click Ask Data at left.
- In the text box reading “Ask about fields in this data source”, enter terms from the data source fields at left.
Hover over each field to learn more about the data it contains. Unlike the data fields seen in standard Tableau views, these fields can't be dragged to the canvas.
- As you type, Ask Data automatically completes your entry using likely combinations of data fields and analytical functions. To automatically create a viz, either press Enter, or choose from suggested entries below the text box.
Just like that, you've created a compelling visualization of your data!
After Ask Data initially builds a viz, you can quickly revise it by typing conversational phrases in the text box reading “Adjust your question or clear all to start over”.
In the text box, do any of the following:
- Change the viz type with phrasing like “as a bar chart”.
- Add functions with phrasing like “by country” or “in December”.
See Supported Analytical Functions for Ask Data for the complete list.
- Replace fields by typing “replace [existing field name] with [new field name]”
- Remove fields by typing “remove [field name]”
- Clear the viz by typing “reset” or “clear”.
Rephrase your question using the interface
You can also rephrase questions by clicking options and data fields in the user interface.
Change the viz type
If the default viz doesn't fully reveal your data, click the menu at upper right, and choose from these supported viz types:
- Gantt Bar
- Heat Map
Note: To automatically create certain viz types, Ask Data sometimes adds fields like “Number of Records” to your entries.
Ask Data gives you several ways to fine-tune how field values are displayed.
- To switch the fields used for the vertical and horizontal axes, click the Swap Axes button above the text box.
- To change a field or its aggregation type (for example, from sum to average), click the field name in the text box.
- If a field is used as a filter, click the name to specify values or wildcard parameters.
- To adjust a numeric range, click words like “high”, “between”, or “cheap” in the text box.
To adjust a relative date filter, click words like “last”, “previous”, or “next”, or “following”.
To delete a field, hover over it and click the X.
Ask Data supports simple calculations between two measures, which you can apply using these symbols:
+ sums the measures
- produces the difference between them
In workbooks you save from Ask Data, these calculations don't become calculated fields but instead ad hoc calculations on the Columns, Rows, or Marks shelves.
Ask Data lets you compare time periods with phrases like “year over year”, “quarter over quarter”, or similar comparisons by month, week, or day. The results appear as difference or percent difference table calculations in workbooks you save from Ask Data.
In the text box, click a difference calculation to choose other fields, aggregation methods, and time periods.
To quickly create multiple different vizzes from a data source, add sheets in Ask Data, and then save them in a new workbook.
At the bottom of the web page, do any of the following:
- Click the Add Sheet icon to the right of named sheets.
- Right-click a sheet name, and choose either Duplicate or Delete.
(To rename sheets from Ask Data, you need to save them in a new workbook.)
You can quickly share Ask Data vizzes without even saving a sheet. Simply click the link icon above the query box, and copy a link viewable by anyone with access to the data source.
After you ask a question with Ask Data, an Embed Code icon appears in the toolbar. Click it to copy code you can incorporate into any web page.
When web pages with Ask Data embedded in them load, Ask Data opens the data source without a viz, waiting for users to ask a question.
To see Ask Data embedded in a web page, users must have a Tableau license and permission to connect to the data source. They can interact with Ask Data just as they would in the Tableau Online or Tableau Server interface, but they can't save resulting vizzes.
Save the vizzes in a new workbook
Save vizzes from Ask Data to combine them with other Tableau views and dashboards in a workbook. In a workbook, Ask Data is unavailable, but you gain access to Tableau's full range of data analysis and visualization tools.
- Above the viz, click the save icon , and choose Save as if you're saving the viz for the first time, or Save to overwrite a previously saved version.
- If you chose Save as, select Embed password for data source if people who lack access will be viewing the viz.
Tip: To quickly navigate to the workbook, click the save icon again, and choose Edit workbook.
Owners of data sources can configure them to be reanalyzed by Ask Data on an automatic, daily schedule, or only upon user request.
If analysis requires user request, and you know a data source has changed recently, click the circled “i” in the Data pane at left, and then click Reindex Data Source.
Tips for success
For example, instead of “I want to see all the countries that these airports are in”, try “by airport and country”.
Use exact wording for attributes and values when possible
For example, if your data source includes Airport Code, Airport Name, and Airport Region fields, specify those by name.
See a ranked list
Ask Data maps terms like “best” and “worst” to Top 1 and Bottom 1, respectively. If you want to see broader rankings, use “high” and “low” instead. For example, enter “houses with low sale prices”.
Surround unusually long values with quotation marks
If you want to analyze long field values that contain line returns, tabs, more than two words, or more than 50 characters, surround them with quotation marks. To improve performance, Ask Data doesn't index fields of that length.