Using the backlog dataset in Insights (Professional and Enterprise)

Zendesk is removing Insights reporting on February 5th, 2021. If you are using Insights, you need to move to Explore, the latest generation of reporting for Zendesk products, before that time. If you need more time to move to Explore, you can apply for an extension using this form before January 22, 2021.

This document shows how to use the Ticket Backlog History dataset specifically within Insights.

First, a little background...

The Ticket Backlog History dataset functions a little differently than most other datasets in Insights. For example, the Tickets dataset will show you the current state of the tickets in your Zendesk, and the Ticket Updates dataset can show past events that are viewable within a ticket's event log.

The Ticket Backlog History dataset will show you a snapshot of unsolved tickets at the end of any given date. Unlike the other datasets, this information cannot be found anywhere else in Zendesk.

Many mysteries of Insights can be revealed by studying the data model. The data model shows the relationship between each dataset and its attributes and facts in Insights. You can find your project's data model by going into Insights and going to Manage -> Model.

If you're not familiar with it, it can look complicated and daunting - there are so many facts and attributes that can be reported on! Fortunately, there is a GoodData Labs app that can make it easier to visualize. After authenticating into your Insights project, follow this link: Data Model Visualizer. You'll see a much simpler graph of all of the datasets in Insights:

You can see that the Ticket Backlog History dataset is isolated from the rest of the datasets - it doesn't have any arrows connected to it. What this means to an Insights report-maker is that you cannot slice or filter backlog metrics by anything other than the backlog attributes (click on the Ticket Backlog History dataset on the data model to view those):

As you can see, it's fairly limited due to the sheer amount of data that needs to be contained in the Backlog dataset. The silver lining is that creating a Backlog report is pretty straightforward!

Cut to the chase, how do I make a Backlog report?

1. Select your metric in the "What" section.
   Currently, the only Backlog metric in a brand new Insights project is "# Backlog Tickets". Of course, you can make your own metrics with the two Backlog facts available - "Backlog Ticket Age" and "Backlog Ticket Count". The Backlog assignee ID fact is used for advanced dashboard filters - most reports won't use this.
2. Select your attributes that you'd like to break your metric down by in the "How" section.
   Remember that you can use only the attributes contained within the Backlog dataset! This includes: Backlog Assignee, Backlog Brand, Backlog Group, Backlog Priority, Backlog Status, Backlog Ticket Age, Backlog Ticket Type, Backlog Via, and Date (Event). No more, no less!
3. (Optional) Select a filter.
   Just like the "How" section, you can only choose attributes that are part of the Backlog dataset.

An example of a typical report can be found in the "Overview" tab of the default Insights dashboards. This report show the # Backlog Tickets metric broken down by Date (Event) and Backlog Status.

An important note about slicing by Date (Event)

If you are slicing the # Backlog Tickets metric by a Date (Event) attribute that is longer than a single day (such as a week, month, quarter, or year), the numbers that are returned will be the sum of the number of backlog tickets for each date in that time period:

This number is not very useful, and most report-makers are looking to find the number of tickets in their backlog at the end of each week, month, quarter, or year. To accomplish this, a special filter needs to be added to the report.

For the sake of brevity, a numeric range filter will need to be added that looks like this:

For those who need more specific instruction on creating the numeric range filter, a short video can be found here.

As you can see, the _spFilter Backlog (Last Day of Month) metric is selecting the last date in each month to use for the month's backlog number. At this time there is not a prebuilt _spFilter metric for week, quarter, or year but it can be easily made by changing the _spFilter Backlog (Last Day of Month) metric to use a different period.

When are the Backlog numbers calculated?

It depends on which Zendesk plan you are on!

• A Zendesk account on the Professional plan will sync its data once per day at the time specified on the Insights dashboard. You can specify which hour this sync happens (in your Zendesk's timezone). The Backlog facts and metrics will reflect the state of the tickets in Zendesk at the time of the daily sync.
• A Zendesk account on the Enterprise plan will sync its data approximately hourly. Backlog facts and metrics will reflect the state of the tickets in Zendesk at the last sync of the day. The exact time of the last sync cannot be predicted and is not necessarily exactly the same each day. Chances are it will be between 23:00 and 23:59 each day depending on available server resources at that time.

Fine Tuning: Learn how to analyze and manage your ticket backlog in Fine Tuning: Analyzing the metrics that matter.