How to know if A/B testing scripts are slowing down your website?

The discussion around whether A/B testing tools may slow down websites due to the additional scripts required to load on a site is ongoing.

Several companies that offer A/B testing argue that their services are extremely fast, causing no negative impact on your load time.

In a world where everyone is claiming that they are the fastest and how do you know?

Asynchronous = Fast?

Their reasoning is largely based on the use of asynchronous tracking code, which they claim minimizes load time interference.

However, it's essential to realize that merely using an asynchronous script does not automatically ensure speed, given that nearly every service uses this same technique.

Content Delivery Network (CDN) = Fast?

Some providers claim that their speed arises from their use of premium Content Delivery Network (CDN) tools. But it's worth taking these claims with a pinch of skepticism, as nothing is preventing other services from adopting superior CDN tools to enhance their speed and performance.

Assessing the speed performance of these services can be challenging due to various inconsistent factors like differing CDN quality, internet speed, environmental conditions, and more. As such, depending on measurements of loading delay in milliseconds to track speed performance can yield inaccurate results.

Smaller File Size = Fast?

What can be more consistently relied on are data regarding the size of the script file and the total duration of the request for that specific script. These factors directly influence the overall speed performance.

It is feasible that any A/B testing service could use a reasonably efficient CDN to enhance their speed. They can also make their scripts asynchronous by using the simple async="true" command. However, the real distinctive factor lies in the uncompressed file size.

Every additional 1 kilobyte of uncompressed file size equates to approximately 1 millisecond of additional loading speed on your website. This effect is even more apparent in a mobile environment.

Final verdict

We can agree that a 17kb script will undoubtedly load faster than a 349kb script, which provides a definitive way to measure the speed of the underlying technology.

For example, consider the following data:

Largest Contentful Paint (LCP) | Uncompressed File Size

Mida - 490ms | 17.2kb
VWO - 725ms | 254kb+
PostHog - 733ms | 115kb+
WebTrend Optimize- 815ms | 170kb
ZOHO PageSense - 839ms | 212kb
CrazyEgg - 810ms | 128kb
Kameleoon - 1060ms | 349kb
ABlyft - 1130ms | 64.2kb
Convert - 1230ms | 203kb
FigPii - 2200ms | 446kb

As suggested here, services with smaller file sizes typically load quicker. Therefore, when assessing speed performance, one should consider the uncompressed file sizes of scripts as one of the key indicator.

Ultimately, there's no way to hide the size of your script, and it serves as a primary demonstration of how efficient the underlying technology is, how fast it operates, and how much it impacts your website when running an experiment.

---

Mida’s impressive script makes it the best choice for growth teams looking for a powerful platform to scale.

To learn more about our platform, book a demo with a member of our team today.