RUM Insights

Table of Contents

Introduction

This page captures a collection of basic visualizations for the data contained in the RUM Archive. It's goal is to allow users to get a feel for the most important insights that can be gained from the data, without having to execute SQL queries themselves.

Note that we intentionally do not include interpretations of the data here, as this can often be quite nuanced. For that, look at our regular blog posts that discuss RUM Insights in depth.

Market Shares

It is important to know which devices, operating systems and browsers your visitors are using to access your sites, as hardware power and feature support can differ widely. This helps you better optimize your sites for your target audience.

Device Type

The following graphs show the usage of different Device Types (RUM Archive currently only tracks 3 device types).

Operating System

The following graphs show the usage of different Operating Systems for Desktop and Mobile devices (OSes with < 1% usage share have been filtered out).

Browser

The following charts show the usage of different Browsers / User Agents for Desktop and Mobile devices (entries with < 1% usage share have been filtered out). The data shown is from March 5th 2024.

Browser Versions

The following chart shows the usage of different versions of popular Browsers / User Agents for Desktop and Mobile devices. Use the dropdown to select a browser and platform. Note that User Agent parsing is done using the ua-parser library.

The data points shown are from the first and third Tuesdays of every month.

While for performance measurements and optimization we often focus on a full page load in its own tab (often as first page in a session) not all navigations are of course like that. In this section, we explore other types of navigations and how often they occur within the RUM Archive dataset.

The data shown is from March 5th 2024.

The following graphs show the usage of different Navigation Types.

  • Navigation is a typical navigation by for example clicking on a link.
  • Reload is when the user manually reloads the page.
  • Back-forward is when the user uses the back and forward buttons to navigate in their history.

Visibility State

The following graphs show the usage of different Visibility States.

  • Visible typically means the user watched the page load happen.
  • Hidden is for example when the user clicks to open the page in a new (yet hidden) tab.
  • Partial is when the page load starts hidden, but the user switches to the still rendering page before it's done (for example, they rapidly switch to a background tab).

Landing Page

The following graphs show how many navigations were to a site's Landing Page.

A landing page is the first page a user visits during a new "session" on a given website. This does not always have to be the "homepage", as it can be any page of a site, as long as it is the first a user visits.

A session is typically (though not always) defined as a series of navigations separate by no more than 30 minutes (after 30 minutes, the session automatically ends and the next navigation starts a new session).

Multi vs Single Page Apps

The following graphs show the amount of (typical) "multi-page app" (MPA) loads versus "single-page app" (SPA) loads. SPA navigations can be broken down into both Hard Navigations (the first navigation to the page) and Soft Navigations (in-page route changes).

About RUM Insights

RUM Insights is intended to grow into a wide collection of default visualizations of the underlying RUM Archive data, updated once a month. The current version is limited to just a few graphs for some key data. If you have ideas about additional graphs or new features, let us know on GitHub.

The BigQuery queries used to generate data for these graphs and their outputs can be found in the separate rum-insights-data GitHub repository.