Regarding website monitoring, 2 approaches coexist – without competing: Real User Monitoring, based on real users browsing, and Synthetic Monitoring that emulates users visits. Let’s focus on the specific benefits brought by this second approach.
Monitor your website performance. Here is an essential requirement as the impacts of the load time are huge and numerous: user experience, bounce & conversion rates or even search engine ranking may be significantly affected. So how to properly monitor your website?
We are quite far away from simply ping servers nowadays. Monitoring tools became much more sophisticated, allowing to gather and oversee tons of data, through 2 monitoring approaches: RUM (Real User or Passive Monitoring) and Synthetic Monitoring (also called Active Monitoring).
Not only these 2 technologies don’t compete themselves, but they are complementary indeed. Good news: you may already have a RUM tool via Google Analytics (and its page speed feature)… So let’s focus on the major differences brought by the synthetic monitoring against RUM and list – at least – 5 cases where this approach is necessary.
Passive vs Active Monitoring
As you could guess, Real User Monitoring is based on data gathered from real users, while they are browsing the website. Unlike Synthetic Monitoring, that emulates users within steady and precise browsing contexts (location, bandwidth, browser, etc.).
As a consequence, those 2 technologies measure the speed of a website in 2 very different ways: while RUM will reveal the complexity related to the users diversity and their browsing conditions (users location, device, connectivity,..), the synthetic monitoring measures the web performance within contexts that are totally under control.
The RUM brings specific benefits that can’t be denied, but also got limitations that can/should be fulfilled by a Synthetic Monitoring approach. Here are a few cases:
#1 Measuring your website performance whatever its stage is.
If users can’t access your website yet (development stage as an example), you just can’t use RUM. Synthetic monitoring is then the only way to measure your web page’s performance!
#2 Get additional & precious metrics
With Synthetic monitoring you are not limited to the data related to users, and you can go much deeper. As an example: Synthetic monitoring is the only way to get a video replay for your tests and to measure the Speed Index. Speed Index is calculated by analyzing the video of the page’s loading.
We have to notice there is now a RUM-SpeedIndex, but the algorithm is still very early stage. It’s not computed from video analysis but from web browser paint events.
Dareboost also provides a Speed Index for your User Journeys (including a video replay!), have a look to our User Journey Performance Monitoring feature.
#3 Compare your website performance against your competitors’
You might have guessed it yet: as synthetic monitoring mimics users’ browsing, you can apply it to whatever website you want, including your competitors! Quite a precious asset, allowing you to launch competitive benchmarks as well as quickly detect any of your competitors attempts to optimize their web performance that may degrade your own position.
#4 Measuring long term results for your optimizations
As synthetic monitoring offers a stable context for your tests, any evolution of your performance will be related to your website evolution. With RUM, you are reliant on the evolution of your users characteristics, and segmenting your traffic (in order to limit that variability) may quickly challenge your results…
#5 Benefiting from alerts to detect any slow down or regression
Your audience may be too limited or heterogeneous to get a stable repository within Real User Monitoring. As explained above, RUM is based on data gathered from your traffic. If this traffic is too limited or heterogeneous (conducting you to segment the data), the resulting data may be too limited to offer you a relevant measurement. Thanks to synthetic monitoring, you will be able to easily set your own thresholds and then benefit from an alerting system! As an example, you will be able to detect any slowdown due to an update.
These 5 concrete cases clearly show the main benefits of synthetic monitoring. By the way, that approach is particularly well adapted to apply the performance budget methodology, and ensure you to offer a fast user experience all along your projects lifetime!