Analyze sessions based on their Web Vitals
You can find and use the page load metrics category when setting your analysis context (device, timeframe and segment) of all analysis modules.
- When setting your analysis context, select the 'condition' drop down.
- Select a metric from the 'Page Load Metrics' dropdown list or search for it by name.
- Set the conditions for your analysis.
- Select 'Apply'.
Note: When analyzing a correlation of web performance and business KPIs, we strongly recommend using the same page view condition on landing or exit page, because our data leads us to think that what matters most is first impressions and the root cause for exiting.
Analyze Web Vitals on user journeys
Using Journey Analysis, find Core Web Vital results displayed on the journey details for each step. Click on them for a shortcut to 'Analyze the web performance' same page in the RUM module.
Learn more about how to analyze sessions with Web Vitals in Journey Analysis
Get proactively alerted to changes in your Web Vitals
Create alerts to notify you (via Slack or email) about changes to your Core Web Vitals. Define your own trigger thresholds with Manual alerts, or let AI alerts detect variations for you.
To create these alerts, follow the steps in this guide.
Alerts can be created for the following Core Web Vitals
Alerts can be configured on the 75th percentile, median or average values for each metric.
- Largest Contentful Paint (LCP)
- Cumulative Layout Shift (CLS)
- First Input Delay (FID)
- First Contentful Paint (FCP)
- Time To First Byte (TTFB)
- Interaction to Next Paint (INP)
Monitor Web Vitals with Dashboards
In Dashboards you can leverage Web Vitals metrics to see their evolution over time on given pages. When adding your widgets to analyze speed related metrics, you will be able to choose from statistical measurements commonly used for web performance:
- Average: the middle value, calculated by dividing the total of all the values by the number of values.
- Median: the middle value, which is found by ordering all data points by their value and using the middle one. For example, 50% of the results are over the median and 50% are under. (Median is also known as the 50th percentile).
- 75th percentile: the 75th percentile is the value at which 25% of the results lie above that value, and 75% of the results lie below it. This measure supports an analysis less influenced by outliers, as it highlights the web performance that the majority of your users experience.
Read more about the 75th percentile
A percentile helps you understand the rank, or position, of a distributed percent in comparison to another distributed percent. For example, by using the 75th percentile in web performance, you might be looking to understand what 75% of your users are experiencing versus the other 25%, relative to a specific value (for example, how quickly the page is loading for them).
Contentsquare example
Let’s use the Largest Contentful Paint (LCP) metric and the 75th percentile measure as an example. Remember, the 75th percentile is showing you the 75% sitting below a specific value (in this example, LCP seconds is our value).
In the line chart below, we can see that on the data point for Sept 10, the 75th percentile was at 0.37s. This means that 75% of pageviews had an LCP of less than 0.37 seconds, meaning that 25% of pageviews had an LCP of more than 0.37 seconds.
Tip: Want to see how many sessions are in that 25%, above the 75th percentile? Create a segment where the LCP (or another metric) is greater than 0.37 seconds (or another 75th percentile value you’re tracking).