Custom Dashboards

In addition to our pre-built dashboards for topics like Crashes, Logs, and Network Requests, each project can create Custom Dashboards. In these dashboards, you can create new analyses. For example, you can visualize a time series of Crash data segmented by app-version, or get a ranking of Logs filtered by a specific property.

To get started, click on "Add new widget" in the menu by your dashboard name.

Creating a Chart

You can create multiple types of visualizations. A table of which visualizations are supported is shown below:

Chart Type	Crashes	Network	Logs	Issues	Moments	Sessions	Spans	Historical	Multi-query
Line	✅	✅	✅	❌	✅	✅	✅	✅	✅
Table	✅	✅	✅	✅	✅	✅	✅	❌	❌
Bar	✅	✅	✅	❌	✅	✅	✅	❌	❌
Pie	✅	✅	✅	❌	✅	✅	✅	❌	❌

Spans (Traces)

Create graphs monitoring Spans performance. You can filter and group by name, outcome, duration, and any custom Attributes you have set on the Span.

• First, select the visualization type you want to see and a source for the query.
• Then, choose the metric you wish to aggregate. For Spans, Embrace supports both counts and sums of duration.
• Finally, add any filters and group-bys. In addition to our core dimensions, for Spans you can use the Span's name, outcome, duration, and any custom Attributes.

Combined Time Series (Line Charts)

You can also create Combined graphs in Line vizualizations. With Combined graphs, you can visualize up to 10 time series together and optionally create a custom time series using arithmetic operators.

NOTE: as of 2025-05-20, the chart builder has been improved. Any legacy charts using fills have been moved to multi-query line charts.

Adding Formulas to Graphs

For example, you want to get a rate over the last 24 hours of the sum of crashes and ANR's with a duration of 3.5 to 10 second relatve to your total sessions, grouped by App Version. To do this:

1. For the inital query A, select Crashes as the source, Crash Count as the metric, and App Version in the group by field.
   a. You can also add a group by here. For this example, we group by App Version. See more on grouping functionality below.
   b. Formulas and multiple queries are only supported on line charts currently.
2. On query A, click the Filter By button below the query details and filter for State = foreground.
3. Create a second query (B) by clicking the "Add query" button. Select ANR's as the source, ANR count as the metric, and again group by App Version.
   a. On query B, filter for Duration Bucket (milliseconds) = 3500 - 10000.
4. Create a third query (C) for your Session Count grouped by App Version and filter for foreground sessions using the same process as in query A.
5. Hide queries A, B, and C by clicking the eye symbols at the right side of each query row.
6. In the Formula field at the bottom, enter (A+B) / C. You can optionally multiply this by 100 to get a y-axis as a percentage (((A+B) / C) * 100).
   a. The Formula field accepts the operators: +, -, *, /.
   b. Order of operations follows standard PEMDAS.

Multiple Combined Time Series without a Formula

Since the Formula field is optional, you can visualize up to 10 time series on the same chart with the steps above. Use the eye symbols to show or hide each, and leave the Formula field blank.

Grouping with Combined Time Series

When you combine time series that have groupings defined with a formula, it’s important to understand the rules for how time series are joined given their groupings. The examples below use two time series, but the same rules apply with any greater number of time series.

As a rule, if you have two timeseries that both have groupings, the grouping of one time series must be a subset of the other. The common subset is the join key. Otherwise the combination is invalid.

Valid Groupings:

• Grouping only on one time series

  • A: grouped by [”App Version”]
    • Returns values for each distinct, non-empty App Version.
  • B: grouped by []
    • Returns 10
  • A has a grouping and B does not. Their common join key is the empty set.

Dimension	A	B	A + B
{"App Version": "1.0"}	100	10	110
{"App Version": "2.0"}	50	10	60

• Grouping on both time series

  • A: grouped by [”App Version”, “Country”]
    • Returns values for each distinct, non-empty combination of grouping dimensions.
  • B: grouped by [”App Version”]
    • Returns values for each distinct, non-empty App Version.
  • Both A and B have groupings, and B is a subset of A. Their common subset is [”App Version”].

Dimension	A	B	A + B
{"App Version": "1.0", "Country": "US"}	100	10	110
{"App Version": "1.0", "Country": "MEX"}	30	10	40
{"App Version": "2.0", "Country": "US"}	50	20	70
{"App Version": "2.0", "Country": "MEX"}	60	20	80

• No groupings

  • A: grouped by []
  • B: grouped by []
  • A and B have no groupings. Their common join key is the empty set.

Invalid Groupings:

• Grouping with no overlapping set

  • A: grouped by [”App Version”]
  • B: grouped by [”OS Version”]
  • Both have groupings but neither is a subset of the other.

• Grouping with similar Attributes, but no overlapping data

  • A: grouped by [”App Version”, “Country”, “OS Version”]
  • B: grouped by [”App Version”, “Country”, “Model”]
  • If A and B have no similar App Version and Country, the grouping is valid, but neither is a subset of the other and no results are returned.

Order of Operations

Additionally, order of operations matters. Consider a situation where you create three queries:

• A grouped by ["App Version", "OS Version"]
• B grouped by ["App Version"]
• C grouped by ["OS Version"]

Then you were to combine them in the formula field:

• Valid: (A + B) + C: A and B share a grouping dimension and can combine, then combine with C.
• Invalid: A + (B + C): B and C do not share a group-by dimension and therefore fail.

Table of Issues

Our Issues Widget, lets you specify how to list a table of Issues . You can filter for certain Issue types, add filters to limit app-versions, or select just Issues tagged to your team.

1. To get started, select the Table visualization option and Issues as the source when making a new Widget.
2. Adjust the columns you want to display and how to filter the Issues.
3. Then once you save, you'll see this table on your dashboard! Issues are sorted by percentage of users impacted in descending order.

Emailing Custom Dashboards

You can also email your Custom Dashboards to yourself or your team.

Click on the "Create Report" button in the top right of your dashboard.

You can set a title for your report, add the recipients, and set the frequency of the report. Frequency options include daily and weekly. You can also set the time of day you'd like to receive the report. You can update any of these options after the report is created.

Before sending the report, you can send a test email to the recipients to ensure the report looks as expected. After a few minutes you should receive the test email.

Grouping by Exploded Properties

You can group your widgets by exploded properties—these are properties that can contain multiple values which we automatically "explode" so each value can be analyzed individually.

This is especially useful if you’re tagging Sessions, Logs, Spans or other events with multiple values and want to break them out into individual groups. For example, if a Session has a property like BRANCH=["master", "develop", "main", "staging"], you can group by BRANCH, and each of those values will be treated as its own group.

To group by an exploded property:

1. When adding or editing a Widget, go to the Group By section.
2. Start typing the name of the property you want to use—exploded properties will appear just like any other attribute.

3. Select the property. If it's a multi-value field, we'll automatically explode it for you.

In your results, each unique value will show as its own row or series—allowing you to analyze each individual value separately.