The Way You’re Measuring Your Web Performance Is Wrong

A man is in the street after dark, searching the ground under a streetlight. A passerby asks what he’s looking for. The man says, “My contact lens, I lost it in that alley.” The passerby is confused, “Why are you looking for it over here then?” The man gestures to the streetlight, “Because the light is better!”  

Everything is wrong about how you measure web performance. You’re stuck obsessing over 2 types of measures: Synthetic and RUM. Both report technical measures such as pagespeed and load, in different ways. But you’re not focusing on meaningful outcomes for digital guests.

The way you're measuring your web performance is wrong. Here's why. Click To Tweet

New web technologies like Progressive Web Apps are changing the way web experiences work, moving away from a page load model. Leading ecommerce sites are moving to PWAs because taking this approach allows the website to react immediately to user input, and fill in required information via round trip to the server later. This places less emphasis on page load, since there’s only one page load that happens (but performance still matters for that first page load!).

Until now, there’s been no emphasis on measuring the qualitative experience, or the time-to-task completion, AKA the customer journey. You need to rethink how you measure web performance, because the way you’re doing it is (mostly) wrong. Focus on the customer task and journey, not the data that the instruments provide.  

ICYMI, “Progressive Web Apps” is a term coined by Google for app-like websites that are reliable, fast, and engaging. Unlike native apps, Progressive Web Apps are accessed through the web browser instead of being downloaded to a device. They work on all browsers, but the functionality is enhanced on browsers that support more advanced underlying technology.

Technical Performance Measures

You’re probably familiar with the key technical performance measures – DOMContentLoaded, PageSpeed, the Load event. These are good for diagnostic and engineering purposes, but they don’t tell the real story of the shopper. True, they are easier to measure, but that doesn’t mean they are worth focusing on, because they don’t represent when an experience is “happening” or “useful.”

Perceived Performance Measures

When it comes to web performance, perception is as good as reality. If a customer believes the experience is fast, it is fast – even if it’s technically slower. Measurements like first contentful paint, first meaningful paint, and time to interactive are a step in the right direction, because they are focused on events that are actually perceptible to the customer. Unfortunately, there are still some big limitations with perceived performance measurements because they emphasize a single page view, or a “quantum of web experience,” as opposed to the holistic customer journey.

Task Completion Measures

Time-to-task completion measures are essential because new technologies like Progressive Web Apps demand a new way of evaluating web performance. The motivation for these these enhanced, app-like web experiences is to keep your customers on your site, get them to purchase something, and keep them coming back for more. That means focusing on performance issues like:

  • Conversion rates
  • Repeat visitors
  • Customer lifetime value

As we reinvent mobile web performance, we need to rethink how we measure success. Time-to-task completion helps us understand the impact of major innovations, like Apple Pay or Android Pay. These quick mobile payments decrease time to complete a purchase from minutes to seconds, yet that vital improvement would not be captured in traditional performance measures.

Still have questions? Get in touch. We’d love to hear from you.

Leave a Reply

Notify of
Resource ad