Impact of embedding Contester solution on PageSpeed Insights analysis results

PageSpeed Insights is a popular tool to evaluate specific webpage performance through lens of end users.

It evaluates complexity and efficiency of various scripts and other resources that are embedded on the website.

We ran their analysis on one of our customer's websites and at most our solution's impact is... negligible.

The following are three reports that we ran within 20 minutes window, though mind that they may have some inherit differences because of customer's website behavior on different country landing pages.

• Page that has no Contester integration
• Page with Contester integration, but not showing our widget
• Page with Contester integration and showing our widget

The loading times are comparable and do not highlight any issues that could be linked back to Contester solution.

It can be summarized like this:

Embedding Contester scripts has none or at most negligible impact on website performance.

Why?

• Every website visitor where our solution is embedded is downloading a tiny 1.5kB script that is additionally cached on Cloudflare and user browsers
• Users download a full solution (anywhere between 500kB to 2MB in size for initial download, depending on configured content) only if their URL contains unique parameters causing our widget to render for them
• Scripts are loaded asynchronously (meaning they do not block website page rendering)
• Scripts loading have short-circuit built-in so even when our platform is unavailable (i.e. during re-deployments) the users don't suffer from it