In Explore, datasets give you access to your Zendesk product data. Each dataset contains metrics and attributes that you use to create reports.
Typically, you'll select one of the default datasets that contains data for the product you want to report on. However, you can also create copies of these datasets for testing (for example, if you want to test custom metrics or apply specific security permissions).
Choosing a dataset
You can choose the dataset you want to use before you create reports, or while you create a report.
Choosing a dataset before you create a report
To get a jump start on creating reports, you can decide which dataset to use before creating a report by going to the Datasets library.
To choose a dataset from the Datasets library
- In Explore, click the Datasets icon ().
- On the Datasets library page, select the product dataset you want to use. The
prebuilt, original dataset names display the status Default, and should be used
If a dataset displays the status Out of date, it is no longer being updated by Zendesk. Unless you are maintaining reports that use this dataset, consider deleting it.
The page for the dataset you chose opens. From here, you can select an existing report or create a new report.
Choosing a dataset when you create a report
When creating a report, the first step is always selecting a dataset.
To choose a dataset while you create a report
- In Explore, click the Reports icon ().
- In the Reports library, click New report.
- On the Select a dataset page, select the Zendesk product you want to create
- Select a specific dataset within that Zendesk product. For example, Zendesk Support
contains different datasets for tickets, updates, SLAs, and backlog history.
When you select a dataset, the Dataset panel appears on the right. This panel allows you to select the Dataset version: either the default version of the dataset (most common) or a custom version. You can also see which metrics the dataset includes, which helps you determine that you're selecting the right dataset before you start creating your report.
- Click Start report.
The report builder opens using the dataset you chose. You can now begin to create your report.
Creating a copy of a dataset
In some circumstances, you might want to create your own copy of one of the default datasets. For example, you might want to experiment with the data structure, or practice working with custom metrics and attributes without worrying about cluttering a shared default dataset.
To create a copy of a default dataset
- In Explore, click the Datasets icon ().
- In the Datasets library, click New dataset.
- On the Select a source page, choose the product you want to create reports for.
- In the Display name field, enter a name for your new dataset.
- Under Select a dataset, select the default dataset that you want to make a copy
- Click Done.
The report builder opens using the dataset you just created. You can now begin to create your report.
Explore offers further options for managing your datasets. The available options depend on the level of permission you have to the dataset. For example, you can't delete the default datasets.
To access the dataset operations
- In Explore, click the Datasets icon ().
- In the Datasets library, hover the mouse over the dataset you want to manage and
click the options menu ().
- From the options menu, select one of the following:
- New report from this: Opens a new report using the selected dataset.
- Edit: Opens the dataset configuration page, where you can change the name of the dataset and review the product it uses.
- Rename: Allows you to change the name of the dataset.
- Clone: Creates a copy of the selected dataset.
- Delete: Deletes the selected dataset.
Any update on the feature to join multiple datasets, or at least the ability to use more than one dataset when building a query? It would be really helpful for us, as we had a few queries in Insights that did not have this restriction, and we can't really replicate them in Explore the way we want to.
Second the ability to join multiple data sets, especially now with the guide data set. We are unable to calculate a self-service score without requestor data. Thanks!
@Spencer, What are you trying to measure in your report. Can you explain more about the self-service score? and requestor data?
Can someone explain to me why anyone would want to duplicate a dataset? Doesn't that open the door for possibly having conflicting reports if a person splinters off the main dataset and starts making their own queries from it?
Does a duplicated dataset continue to have the same new data as the original it was copied from?
Hi James McCoy, thanks for the question. I think the main use for cloning a dataset is for testing purposes, for instance if someone wants to create a bunch of custom metrics and attributes that might not be ready for prime-time yet. If you restrict other agents from using that dataset that you can continue to develop until you have your queries correct. You're correct however, you don't want people getting confused about which dataset is which.
I see in the comments above that 4 years ago it was in the product queue to allow queries across data sets. Did this ever happen? I need to be able to create a report that totals up customer interactions (i.e inbound calls, outbound calls, chats, inbound:outbound emails, web form submissions) and while I can do each of these with individual queries and put them on a dashboard, I see no way to total them without exporting the data and manually tallying the results.
+1 for needing the ability to create queries across datasets!
We are implementing Chat on a phone-only based team, so I want to be able to track the number of inbound calls and number of chats served for that specific group as one metric.
If you're looking to allow the creation of queries across datasets, please upvote and add your use cases to the following product feedback post – thanks! Create Query from two different datasets
We also need the ability to query across Datasets. We often run into issues with being unable to adequately query due to the datasets being separate. Currently, I'm trying to find the Time Spent on a particular type of ticket, but I can't isolate the type of ticket in the updates dataset and I can't view the time spent in the support dataset.
I'm trying to do something which would presumably be very simple, but I don't think it is possible. I am trying to build a single report/pie chart that shows volumes of inbound contacts by channel (in our case voice, email, and chat). Doing this using the 'Support - Tickets' dataset leaves out any calls that didn't create a ticket (disconnected/abandoned calls... and though we can change the setting to create tickets for every abandoned call, that seems like a poor and unnecessary option to enable better reporting).
We're also not able to filter these contacts by relevant metadata (e.g. 'Call direction' in order to omit outbound contacts on voice) using only 'Support - Tickets' available data.
That's right, that is currently not possible as the data that you need to report on are in separate datasets. If you can, I recommend that you also upvote and post to the product feedback thread that Dave mentioned above – Create Query from two different datasets. Thanks Robert!
Is there a way to get an export of the out of box datasets, or can you only export reports?
I need to be able to query against two datasets (I've already upvoted the link mentioned in this thread), and also want to be able to use other visualization tools, including to conditionally highlight (https://support.zendesk.com/hc/en-us/articles/4408819953818/comments/5003965656474). I'll ultimately likely needs to display things in an external dashboard that draws data from other systems, too.
I'm afraid there isn't a way to export the datasets directly from Explore. You can try exporting via the API (see API Reference; Incremental Exports); essentially, the data available in Explore are pulled from these endpoints.
Hi! Is there any way that I can create a custom dataset, filter it to a particular group, and data showing only that filtered group can be seen by people I share it with (so they can build reports for it but not see other groups' information).
Unfortunately, this is not possible in Explore. When cloning datasets, all metrics and attributes (including custom ones) are copied. There's no option to exclude some elements from being cloned, or limit access to data of certain users within a dataset.
Please sign in to leave a comment.