Plan requirement
| Subscription | Suite Professional or higher, Explore Professional or higher |
| Access | Admin |
Choosing one, copying one, and the status on a dataset that tells you to delete it. Leave the standard datasets themselves alone.
Work with them
- Click the Datasets icon in Explore to open the library, or pick the dataset while building: Reports, New report, then Select a dataset.
- Choose the product, then the dataset within it. Support alone has separate datasets for tickets, updates, SLAs and backlog history, and choosing the wrong one is why a metric you expected is not there.
- In the Dataset panel on the right, check the Dataset version, default or custom, and read which metrics it includes before you start.
- Click Start report.
Read the status on each dataset
A dataset marked Default is the prebuilt original, and Zendesk's advice is to use those wherever you can.
One marked Out of date is a legacy dataset Zendesk no longer updates. Some of its metrics and attributes may not work correctly, and unless you are keeping old reports alive on it, it should be deleted. A report built on one of these looks entirely normal and can be quietly wrong, which makes it worth a look through your list today.
Clone rather than edit
Hover a dataset in the library, open the options menu and select Clone. Whether other people may do that is an admin setting, the Clone dataset permission.
The reason not to edit a standard dataset is that it is shared by every report using it, including the prebuilt dashboards. Adding to one affects all of them, and removing something breaks reports you did not know existed. A copy is also the place to test a custom metric, or to apply narrower security permissions than the original carries.
Why you would clone at all
To add calculations your organisation needs everywhere: a grouping of ticket types into your own categories, a flag for your own definition of an escalation. On the dataset, every report can use them, instead of the same formula being retyped in twelve places.
Keep the number small, and name them properly
One custom dataset per real need. Five near-identical clones produce reports that disagree and nobody can say which is authoritative. Put the difference in the name: "Tickets with product grouping" is useful, "Tickets copy 2" guarantees the wrong one gets used.
Document the calculations
A calculated field is a decision about how your business defines something. Write down what it means and why, because in a year the formula will still be readable and the reasoning will not.
A new calculation is not visible immediately
Datasets fill on the refresh schedule, which is daily or hourly depending on your plan, so a calculation you just saved shows an empty column until the next sync. That is timing rather than a mistake.
Comments
0 comments
Article is closed for comments.