Skip to content

Thinking beyond Largest Contentful Paint, RUM best practices & case studies from Etsy and GOV.UK

Sep 1, 2022


Hi there,

Is it just me, or does it feel like there's a lot going on in the world of web performance?

Maybe it's the return to in-person conferences and events (see below for more on that). Or maybe it's the case studies being shared by performance leaders like Etsy and GOV.UK (also below). Whatever the cause, I'm here for it!

Without any further ado, let's get into it. As always, I'd love your feedback and suggestions for making these newsletters helpful. Just hit the 'reply' button to get back to me.

Until next month, I hope you stay safe, happy, and well.



US Retail Benchmarks: Kohl's takes the lead

Fall is a huge time of year for retail, so site speed and user experience matter more than ever. The latest Industry Benchmarks snapshot for US retail sites shows Kohl's in the lead with a Largest Contentful Paint (LCP) time of just 0.9 seconds on desktop!

What makes Kohl's fast? Looking at their latest test details, a few things jump out:

  • Speedy backend time (0.21s)
  • Short and simple rendering path
  • Excellent resource optimization

Thinking beyond Largest Contentful Paint

Default metrics – like Core Web Vitals – can be a great starting point if you're just beginning to measure the performance of your pages. But if you want to home in on measuring what matters most on your pages, eventually you're going to want to dig deeper.

For example, Largest Contentful Paint (LCP) is a Core Web Vital that gets a lot of attention, but it comes with a handful of caveats. In this post, Andy Davies (@AndyDavies) explains:

  • The pitfalls of LCP
  • Why Element Timing is a better way to track the performance of the content that matters most on your pages
  • How Element Timing is better for SPAs

Priority hints: What your browser doesn't know yet

How does Etsy keep up with advances in web performance? In this recent blog post, David Ross (@cdaveross) walks through Priority Hints and the road to testing faster page loads.

GOV.UK sees 17% improvement in interaction metrics after removing jQuery

Longtime SpeedCurve RUM users GOV.UK recently shared a project where they dropped jQuery from their pages and scored some major UX wins, including:

  • 17% improvement in Visually Complete
  • 17% faster Time to Interactive
  • 9% faster Load Time

Learn more about how they did it and the results they saw.

Web performance is for everyone

Photo by Shane Rounce on Unsplash

If you're a developer or engineer, there's a good chance you've been doing performance work for a while. Over the past couple of years, I've noticed a lot of people coming to performance from other fields, which is so great to see. If you're one of those people – or if you know someone who is – I recommend checking out and sharing these guides:

Is there a guide you'd like to see? Let me know!

The right way to sample your real user data

"Can I sample RUM data? If so, what should my sample rate be?"

I hear this question a lot. The good news is that, yes, you can sample your RUM data and get meaningful insights from it – while at the same time managing your monitoring budget. But there's a right way and wrong way to do it. Cliff Crocker (@cliffcrocker) analyzed real user data to arrive at these best practices for calculating the best RUM sample rate for your site.

Remember to self-host your static assets

Further to the tweet at the top of this email, Harry Roberts (@csswizardry) reminds us:

"One of the quickest wins—and one of the first things I recommend my clients do—to make websites faster can at first seem counter-intuitive: you should self-host all of your static assets, forgoing others’ CDNs/infrastructure. In this short and hopefully very straightforward post, I want to outline the disadvantages of hosting your static assets ‘off-site’, and the overwhelming benefits of hosting them on your own origin." returns to Amsterdam!

We're thrilled about the return to in-person conferences, and we're especially thrilled to be a sponsor of in Amsterdam on October 27-28. There are two SpeedCurvers on the stage this year:

  • Andy Davies (@AndyDavies) will be sharing approaches and practical steps he uses to help people reduce the impact of third-party tags on the speed of visitors' experience.
  • I'll be revisiting old assumptions about performance budgets and introducing some new practices to consider.

The speaker line-up is absolutely stellar – Katie Sylor-Miller, Alex Russell, Nic Jansma, Katie Hempenius, Stoyan Stefanov, Harry Roberts... I could go on and on, but you get the idea. If you're going to be there and would like to connect, let me know!