User centric performance metrics - Lab vs. RUM
The importance of web performance is a standard best practice. A fast website is better than a slow one. But what is fast? As Philip Walton points out, web performance is relative.
A website might be fast for one user, but slow for another. Your website appears to load fast but can’t respond to user interaction. The relativeness of performance illustrates the need for objective and measurable data.
You can measure performance data (metrics) in two ways:
- In the lab
- In the field
In this post, I show you two approaches to measure the performance metrics or your website. I outline the use-cases, advantages, and disadvantages of lab and field measurements.
Measurement approaches - lab and field data
Measuring performance is essential for your website’s health. There are two common approaches to measure performance in the performance world. Lab-based testing and gathering field data, also called Real User Monitoring (RUM).
Measuring the user experience requires more than one single Lighthouse test. It is also not enough to compare one single metric. User experience is dependent on the user’s journey through your website. You need to measure this journey throught different metrics.
Hey friend, you are about to get my free newsletter straight into your mailbox. I mainly write about web performance tips and trends. Occasionally, you get an email about my workshops, privacy, and web accessibility.
Furthermore, a single test on Lighthouse doesn’t represent your users. In the real world, network conditions and devices vary a lot. Some users might visit your website from a fast network, while others are on a slow mobile connection. These facts clarify that reliable web performance monitoring needs lab and field data.
Lab data is performance data from a controlled environment. You predefine the devices and the network settings adjusted to the users that are on the website. Lab data is an excellent choice to reproduce bugs, and fix performance issues.
Tools like PageSpeed Insights, WebPageTest, or DebugBear can collect and analize the lab performance data.
Advatages lab data
- Easy to set up
- Excellent for debugging, isolating, fixing performance issues
- Fast start into web performance
Disadvatages lab data
- Doesn’t measure the real user performance
- Can not correlate to business values
Field data - Real User Monitoring
The performance of a website can vary based on the user’s device or the network connection. Field data (Real User Monitoring) is performance data from real users on your website.
The device and network conditions are not the only reason why RUM is necessary for a reliable web performance concept. The display of personalized content can take a different amount of time. Lad-based testing isn’t able to measure that.
Field data is expensive to set up unless you use some custom analytics tracking like Plausible. There are many useful tools like SpeedCurve that offer you the possibility to see how real people experience your website’s speed.
Advantages field data
- Alerts and regression prevention
- Real-world examples of the page
Disadvantages field data
- Often expensive
- Third-party integration
In this post, I showed you two approaches to measure the performance metrics or your website. Additionally, I outlined the use-cases, advantages, and disadvantages of lab and field measurements.
If you like this article, smile for a moment, share it, follow me, check out my RSS feed, and subscribe to my newsletter.