Within Power BI, many times users need to perform ad-hoc exploration of their data. This could be an analyst who just got access to a new dataset or data source and wants to spend time learning about the data before building a report off it. Or this could be a business user who needs to answer a specific question using the data to include in a PowerPoint presentation, but the report they’re using doesn’t answer the exact question they have. Creating a new report from scratch in these cases is a large hurdle, just to get a quick answer or screenshot for a deck.
Introducing the public preview of the new Explore feature, where users have a lightweight and focused experience to explore their data. Similar to exporting and building a PivotTable in Excel, now, directly within Power BI users can quickly launch Explore to begin creating a matrix/visual pair to get the answers they need without all the distractions and extra complexity of reports.
- Simply find a dataset or report you’d like to explore
- Begin exploring by building your matrix or PivotTable by adding fields from the data pane
- Add filters if needed
- Save and share your exploration
Step 1: Find a dataset or report you’d like to explore
For a given dataset, report, or Datamart you can begin your exploration by using the “…” menu option from a workspace list or the data hub list view.
You’ll also see this option in the Data hub details page as part of the action card:
Or you may want to explore the underlying data behind a report. You can also begin exploration from the report’s”…” menu while in a workspace list or the report’s menu bar.
Note: Explore is not currently supported for streaming datasets.
What if while working with a Datamart you’d like to visualize and explore the subset of the data produced from an ad-hoc query you just ran. Now, we also provide you the option to use the new Explore feature to dig into your query results as well.
Note: There is currently a known issue when attempting to save an exploration from the Datamart SQL editor entry point. We’re working on a fix.
Step 2: Begin exploring by building your matrix or visual
Regardless of where you start from, launching the explore experience will land you in the new focused view. Your data will be on the right, and you can edit drag and drop or check fields to start building out a matrix or visual.
From a dataset, the experience starts with a matrix, as we know users often start exploring by building out a matrix or PivotTable to better understand what’s available or get down to the right subset of data before visualizing. Whereas from Datamarts, as you’re coming from a grid view of the data already, we start you off with a visual instead.
You can expand the visual representation of the matrix on the lower part of the dialog. This visual maps to the same data as what is shown in the matrix and as you modify what fields are being explored, it will automatically update and select the best visual for that set of fields.
If you prefer, you can swap the layout to show the visual and matrix side-by-side instead of one on top of the other using the layout buttons, , in the top right to adjust the orientation.
If you want to use the field well to adjust which fields are located on the rows and columns, you can use the ‘Rearrange data’ section of the data pane. You’ll only see the field wells of one visual, either the chart or the matrix.
To see the other field wells, you’ll just need to click on visual you’ll want to edit. You’ll know which one you’re editing by the black outline.
Just as when you were editing the matrix and the visual updated automatically, as you edit the visual, your matrix updates to reflect the same data. This applies not only to the fields used, but also interactions like sorting and drilling. Assuming the visual can accurately reflect the sort or drill state, it will update appropriately.
For example, here I’m using a date hierarchy and want to drill down to quarter.
When I do so on the visual, the matrix also updates accordingly.
Note: There are certain matrix configurations that cannot be currently visualized in other chart types. Depending on the situation, a visual may not be rendered or the sort and drill state may not apply. For example, if you expand one specific row in the matrix, that will not apply in the visual, as we don’t currently have support for that interaction within other visual types.
Step 3: Add filters if needed
As you build out your exploration, you can also take advantage of a new filtering experience. Just use the Add filter button on the top left to quickly add filters on the fields already in your visual or pick from any fields in your dataset.
In this exploration experience, the filters are added as pills at the top of view, so you can easily see what the current filter state is and quickly modify it on the fly.
Step 4: Save and share your exploration
To save and share your exploration, click the Save button in the top left corner.
And you’ll get a dialog asking you to name the exploration, pick a workspace to save it to, and add a sensitivity label if applicable.
Note: Explorations can only be saved to premium workspaces at this time.
Now you’re ready to share your exploration with others. Simply use the share button in the top left.
Choose your audience and permissions by clicking the first section and Apply.
Then continue to add your recipient’s email address and share. Immediately following you will also be prompted to share the underlying dataset so the recipient can view the contents of your exploration.
Note: Usually for reports sharing the underlying dataset is done behind the scenes, we’re working to remove this extra step.
You may want to save your exploration as a report instead. You can do so by choosing Save as report from the save button. This will save your exploration as a single page report matching the layout when saved. For example, if you are viewing both the matrix and visual, you’ll get one page with a matrix laid out on the top half and the visual laid out on the bottom half of the page.
You can continue to share this new report as with other reports by using the share button.
- Exploration authors need build permissions on the dataset to create an exploration (open the explore dialog).
- When saving, the destination workspace must be in a premium capacity.
- If saving to “My workspace” you may need to refresh your page to see the newly saved exploration artifact.
- Changing a sensitivity label on a dataset does not currently propagate down to the exploration it was created on.
- There is currently a known issue when attempting to save an exploration from the Datamart SQL editor entry point. We’re working on a fix.
- Explore is not currently supported for streaming datasets.
- Exploration users need a Pro/PPU license to share.
- When sharing, the exploration author will also need to grant build permissions to the underlying dataset so the recipient can view the exploration. If the author skips this second step, the recipient will be unable to see the exploration and have to request permissions.
- Explorations are not currently shown in lineage view as related to the dataset.
- Explore is not currently available from datasets and reports on PBI Home or the Create page.
- All native visuals are supported with the exception of: Matrix, Q&A, Smart Narratives, Metrics, Paginated Reports, Power Apps, Power Automate.
- Custom visuals support coming soon.
Please try out this public preview and let us know what you think! This is just our first milestone of a much larger roadmap, so expect things to change and improve during the public preview period. We want to make sure that this feature evolves to meet your needs, so please share any feedback you have. We are interested in any and all feedback you have, but there’s a few areas we’re particularly interested in:
- Examples of how you or others in your company would want to use this experience
- How useful are you finding the feature? What could we add to make it more useful?
- Do you see yourself coming back to a previous explore session and what your use cases for that would be?
- What are your use cases for sharing out explore sessions (if any)?
- Are there specific visual or matrix features or interactions you’d like to see in this experience?
Drop your feedback into our survey: https://aka.ms/ExploreSurvey
Thank you in advance for taking time to try out our new Exploration feature and providing feedback!