What is an API error?
API errors happen when something goes wrong with an API request, like if there is a missing parameter or a typo, for example. Some of these errors can impact user experience while others go undetected by users. Monitoring API errors helps to ensure a drop in your site's performance is not due to an issue with an unresponsive or errorful API request.
Example: If a payment provider API is not responding due to a typo in the request—then a user will be unable to complete their purchase; or if an inventory API is not showing correct data because a parameter is missing, users might be able to purchase an unavailable product; and so on.
Setting up your Error Analysis scope
1. Set your Analysis context at the top of the page to view errors for selected devices, time periods, and segments.
2. View the list of sessions impacted by errors. In this table you can see the list of grouped errors that occurred within the scope of your Analysis Context.
3. Search for a specific error or click on a different tab to filter by Errors by Screen or Errors by Screen Group.
View and investigate errors from a global list
You can hover over the error details to view the technical details of the error:
- Request URL, method and status code for API errors
- Error text
Investigate the details of the error using the '>' arrow to open the side panel.
In the Error side panel you can complete the following actions:
- Dismiss the error (admin or expert user roles only)
- See the general error details
- Copy error details
- Investigate the impact of the error with a direct link to Impact Quantification and open the embedded player to view the session where the user experienced the error.
View and investigate errors from screens
Use the ‘Errors by Screen’ tab to identify top errors occurring on screens and dig into specific KPIs to understand the impact of the error.
1. Click on the 'Errors by Screens' tab
2. Click on ‘View ___ errors’ of the error row you want to investigate further.
3. You will see a list of errors that occurred on that screen. These errors are prioritized based on the missed opportunity associated with the conversion goal and its statistical significance.
4. Investigate the details of the error using the ‘>' arrow to open the side panel.
View and investigate errors from Screen Groups
1. Click on the 'Errors by Screen Groups' tab
2. Choose the relevant mapping to see which errors are impacting users who reached that specific screen group
3. Identify the page group with the highest lost conversions (using the values in the lost conversions sum column) and click 'View ___ errors' under the screen group name.
4. Investigate the details of the error using the ‘>' arrow to open the side panel.
5. You can easily shortcut to Session Replay, Impact Quantification and Journey Analysis.
API calls we collect
In Error analysis (Apps) API error monitoring tracks two types of HTTP-based APIs, sent by the user-agent (browser) to Servers:
- XMLHttpRequest API: The most commonly used API, which can be used to exchange data between a user-agent (browser) and a server.
- Modern Fetch API: Similar to XMLHttpRequest APIs but with a more powerful and flexible feature set.
Investigate the status code
By reading status codes, developers can quickly see what is happening with the application and quickly check for errors.
There are two main categories of errors:
- Client errors --- status 400 to 4xx
- The error comes from the initial request sent by the browser/web-app, meaning it contains a mistake.
- These errors are more problematic, as the error comes from the code itself in most cases, and reloading the page or resending the request won't solve the problem.
- Server errors --- status 500 to 5xx
- The error is fired during the process of the request by the server, meaning the server has encountered an internal error and can't resolve the request.
- Most server-side errors are temporary, if the user-agent sends the same request later on, reloading the page, the request might succeed.