Finding Web Performance Issues Faster

Eroan Boyer is a web performance expert and founder of Agence Web Performance, a specialist agency in France. His team audits websites, modifies code, and monitors ongoing performance across every major CMS—WordPress, PrestaShop, Shopify, and more. Performance isn't a project they hand off. It's an ongoing service.

For years, Eroan relied on synthetic tools: WebPageTest, Lighthouse. They were the industry standard. But over time, he saw their limits clearly.

Eroan Boyer

Founder, Agence Web Performance

"There is no choice—we have to use real user monitoring data. It's the only tool that can offer real world accuracy. Synthetic tools can be useful in some cases, but if you really want to follow the evolution of a website over time and help clients improve, you have to use RUM."

The Chrome User Experience Report offers a slice of real-world data, but its browser coverage is too limited to rely on. To understand what users are actually experiencing, you need instrumentation running on your own site. That realization is what led Eroan to Request Metrics.

10 Minutes to a Root Cause

Eroan ran a one-month trial alongside his existing tool. He started digging into the data immediately. "The first 10 minutes I opened the tool, I found a CLS issue," he says. By filtering through the interface, he identified exactly which group of pages had the problem—and what was causing it.

That speed matters for an agency. Every hour spent hunting a root cause is an hour not spent fixing it or moving to the next client.

A TTFB Spike That Came From an Unexpected Place

One of Eroan's clients was running Facebook advertising campaigns. Traffic was good, but something was wrong: TTFB on certain pages had spiked to five and six seconds. The culprit wasn't obvious.

In Request Metrics, Eroan pulled up the slowest pages and started filtering. The pattern emerged quickly: every slow page shared the same URL parameters—the click-tracking parameters Facebook appends to ad URLs. Those parameters weren't handled by the client's caching layer, so each ad click was generating an uncached request. The fix was targeted and fast.

Eroan Boyer

Founder, Agence Web Performance

"We can go from the overall dashboard view to pinpointing a root cause—and that information is really valuable. We can tell our clients exactly what we found and how fast we can fix it."

More Than Core Web Vitals

What kept Eroan using Request Metrics beyond the trial wasn't just speed. It was depth. Core Web Vitals are essential—but they're not the whole picture. His agency needed to see JavaScript errors, understand how users navigate, and drill into technical data beyond the top-line CWV scores.

"As web performance experts, we sometimes need to take a step back and see a website from a quality perspective," Eroan explains. "Request Metrics allows us to do that—and we didn't have that previously."

The agency also used it to surface an image issue that wouldn't have been obvious otherwise: file names with inconsistent casing were causing images to fail to load correctly. Once Request Metrics surfaced the heaviest offenders, Eroan's team had a precise list to send to the client. The issue was resolved in under a week.

The Last Tool in the Stack

Agence Web Performance recently expanded into hosting, maintenance, and monitoring—a premium managed service for e-commerce and SaaS clients who want a single point of contact for everything performance-related. Request Metrics is the last tool in that stack.

"Request Metrics is the final step of everything we do," Eroan says. It's how his team measures the effect of their work—and how they prove that value to clients. Every week, the team opens the dashboard, checks that Core Web Vitals are stable, and looks for anything that's shifted. When something has, they have the data to explain it.