Once you have configured RUM, click Reports>Real User Monitoring, locate the domain your RUM monitors, and click on its name from the RUM Check column to access your RUM data.
This section will discuss parsing data by aggregation and will briefly cover reporting by URL.
To skip to a specific section of the "RUM Report" video, click the links with the below!
- RUM Dashboard
- Aggregating, Reporting on, and Working With RUM Data
- RUM Aggregation Toolbar
- Examining Data By Segment
- Select or Deselect a Data Point
The RUM Dashboard provides a snapshot of RUM data with the goal of answering: what is the current state of my application? Is there a problem with my site’s performance?
The page contains several columns that compare specific values over the past 30 minutes of data against the previous 1-week’s baseline.
The RUM Check column contains the name of your RUM monitor(s), the main domain name(s) being monitored, and a Red, Yellow, or Green indicator of the check’s performance. This color coding is based on the Apdex threshold for the check.
The Error Rate column shows the Error type that has the highest change over the previous 30 minutes.
The baseline for Time to Interactive (TTI) is the median of that data point over the previous one week period.
Once you have clicked on RUM domain you have multiple options for viewing RUM report data that will determine what you see, and how graphs appear. Beginning with the RUM Aggregation Toolbar, adjust the date range, aggregation method, or URL/URL group to explore your data.
Use the RUM aggregation toolbar to set the data aggregation method according to URL, date range, or choose an aggregation function for load time values; 95th percentile, 75th percentile, etc.. By default, RUM reports use the last 30 minutes as the date range.
Begin typing the URL pattern, or the URL itself, and Uptime.com will populate a list of similarly matching URLs. Select a URL or group to view all metrics specifically generated by that URL or group.
It is possible to adjust the date range to reflect the previous 1 hour, 3 hours, 12 hours, 24 hours, today’s data, yesterday’s data, the past 7 or 30 days, this month, last month, this year, last year, or by a custom date range.
Date range is an important filter, as it can affect the visibility of certain datapoints. A good example is bounce rate, where the default date range would typically show as 0% with the default date range of 30 minutes. 30 minutes is a good measure of what is happening now in a general sense, but the past hour or even the previous 3 hours will reveal more data.
You can select an alternative aggregation so that your reporting examines:
- Average - mean of all page views
- Median (represents 50th Percentile) - ½ visitors < value of the median, ½ is greater
- 75th Percentile - 75% of page views
- 90th Percentile - 90% of page views
- 95th Percentile - 95% of page views
- 98th Percentile - 98% of page views
- 99th Percentile - 99% of page views
Note: the above percentiles for TTI represent the TTI value that the percentile (e.g. 99%, 98%, etc.) are faster than.
If you want a more detailed understanding of the experience of most users, you might consider adjusting to the 95th or 99th percentile, versus selecting average which is usually easier for high-level reporting.
You can compare what those graphs might look like, and how they might differ with the example below:
The average data set shows a lower Time To Interactive (TTI), First paint, and overall a lower baseline.
Aggregating by the 95th percentile shows a higher TTI and higher overall baseline. We can also see a greater fluctuation in TTI than the flatter graph we see in the average data points.
When viewing the 95th percentile, we see in the above example that a significant amount of users are experiencing some performance hiccups. We would typically suggest the load time distribution histogram as your next best indication of how the actual user population experiences performance.
Auto Refresh and Full Screen View
RUM also has an Auto refresh interval, which can be set from 2, 5, 10, 15, 30, and 60 minutes, or the option to Refresh Now. Additionally, you can use the fullscreen button (useful for portrait display in offices or on a second screen).
A RUM report contains multiple data points broken down by:
- Page Load Time | Skip to 3:09
- Ajax Load Time | Skip to 7:52
- Errors | Skip to 10:01
- Bounce Rate | Skip to 10:44
The default tab is Page Load Time.
Each of these tabs contains a by Segment section, which is further broken down based on filters applied.
Data by URL
Breakdown data by URLs to see specific RUM metrics for a URL or pattern-matched URL group you are monitoring. This option gives URL specific data, so it can be a useful means of focusing on URLs that have high traffic, a growing percentage of errors, or for URL groups.
Data by Device
View a breakdown of data comparing mobile versus desktop access, which provides a direct comparison of page views, showing volume of views from mobile and desktop. For further details about operating system, we suggest breaking down data by Platform. It is also possible to view data by browser, to help identify which browsers are performing best.
Data by Platform
Data by platform allows you to observe how various operating systems are able to access your content. Android, Windows, iOS, Linux and MacOS are possible data points in this graph.
You can also open individual URLs in a new tab when you click the Open in a new tab button.
Data by Browsers
Filter data by the browser type used to access your content. If a browser is not a recognized user-agent, data will be displayed as “Unknown”. Useful for identifying both the number of pageviews by browser, as well as TTI and other key performance metrics for a direct comparison of performance browser to browser.
Data by Countries
Filter by geographic location, arranged by highest sessions.
You can alter graphs to display a specific data series. Use the Legend of each graph to click a data point you wish to exclude and it will be filtered from results.
You can also hover over a specific data point to see it highlighted in your report graph.
In this example, First byte is our highlighted data point, and appears in darker contrast to other data points we wish to filter.
Want to see our checks in action? Check out our YouTube Library for more!