Error Grouping

In Error Analysis, identical errors are grouped to avoid their duplication in the error list, as although multiple errors might be identical, they can have slightly different details (depending on the error type) which could result in them being treated as separate, “unrelated” errors.

Therefore, grouping logic is applied to enable you to:

• Navigate the error module more easily
• Spot and prioritize errors faster
• Identify the real impact of the error

How grouping works by error type

For each of the error types, the following error details are used to group identical errors:

API errors

• Request URL
• Status code
• Method
• Matching body content (if set in a collection rule)

Console messages

• Message, but ignoring values:
  • Message, but ignoring values:
  • Date/Time
  • Numbers
  • Amount
  • Telephone
  • ID
  • JSON values
  • Duration (e.g. 03:35)
  • Identifiers in the form of hexadecimal numbers

Example 1
The following errors would be grouped, as the date/timestamp pattern is ignored for grouping:

• [ERROR]: 2024-06-27T14:14:36.329Z Something went wrong
• [ERROR]: 2024-06-27T13:11:35.391Z Something went wrong

Example 2

The following errors would be grouped, as the number pattern is ignored for grouping:

• [ERROR] This date does not correspond to the age entered in your search (10 years old). Check the date or start a new search
• [ERROR] This date does not correspond to the age entered in your search (25 years old). Check the date or start a new search.

Custom error

• Message, but ignoring values:
  • Date/Time
  • Numbers
  • Amount
  • Telephone
  • ID
  • JSON values
  • Duration (e.g. 03:35)
  • Identifiers in the form of hexadecimal numbers

Example

• “Oops, discount code invalid” or,
• “One or more of the values you have entered is incorrect. Please verify your login information and try again.”

Crashes

For Android Native crashes, we use the following elements of the first frame:

• Exception type
• Package
• Method name
• File name

For iOS Native crashes, we use the following elements of the first frame:

• Exception type (or Signal if no exception type)
• Module
• Function name
• File name

For React Native crashes:

• Context line function name
• File name

Flutter

• Exception Type
• File name
• Class name
• Method
• Directory

JS errors

• File name domain but ignoring values related to country and language (extension codes…)
• Error message but ignoring some values (numbers, URLs) and eliminating specificities unique to browsers or browser versions

Analyzing error groups with a Group ID

Errors that have been grouped together are given a Group ID, so that it’s easy to identify which group an error belongs to. You can find and use error Group IDs when analyzing in Error Analysis, Session Replay and with segments in the Analysis Context.

Error grouping in Error Analysis

Click on an error from the error list to open the side panel. There, you can see the Group ID and click the three-dot ‘more’ menu to “Copy error details”.

Error grouping in Session Replay

When playing back an error via Session Replay, the list of replays is composed of every replay where errors belong to the same error group.

In the Event Stream on the left, view the error's name, details and the Group ID the error belongs to.

Error grouping with segments in the Analysis Context

Create and save a segment of the grouped error in the Analysis Context to use in other modules.

Tip: Use saved segments in Impact Quantification to compare users who have experienced the error vs. users who have not experienced the error, to see the impact on conversion and revenue.