LifterLMS Backup and Disaster Recovery: What to Save and How
What a LifterLMS backup must include, how often to take it, and how to restore without losing orders or double-charging students: a data map of…
Read articleCore Web Vitals are Google's three measurements of real-world page experience: how quickly the main content appears, how quickly the page responds to a tap or click, and how much the layout jumps while loading. On a LifterLMS site they matter twice over. Course sales pages compete in search, and logged-in learners judge your product by how lessons feel on a phone.
This guide is for site owners and developers who have seen poor scores in Search Console or PageSpeed Insights and want to fix them properly. It covers which pages to measure, the causes of poor LifterLMS Core Web Vitals that we see most often, and a short plugin that removes two common bottlenecks.
For general speed work, such as hosting, caching layers and image optimisation, start with our guide on how to speed up a LifterLMS site. This article focuses on the three metrics.
| Metric | Measures | Good | Typical LifterLMS culprit |
|---|---|---|---|
| Largest Contentful Paint (LCP) | When the largest image or text block appears | 2.5 seconds or less | Slow server response on uncached pages, hero images, video embeds |
| Interaction to Next Paint (INP) | How quickly the page responds to input | 200 milliseconds or less | Too much JavaScript: page builders, tracking tags, sitewide plugin scripts |
| Cumulative Layout Shift (CLS) | How much content moves unexpectedly | 0.1 or less | Late-loading banners, fonts, and elements resized by JavaScript |
Google assesses each metric at the 75th percentile of real visits, separately for mobile and desktop, as explained on web.dev's Web Vitals overview. A page passes when three out of four visits are good, which is why a fast desktop test can hide a poor mobile reality.
A LifterLMS site has very different page types, and each behaves differently:
Use field data first. Search Console's Core Web Vitals report groups similar public URLs and shows which groups fail. PageSpeed Insights shows field data for a URL or the whole origin when enough Chrome users have visited, plus a lab test. For logged-in pages, field tools see little, so record a performance trace in Chrome DevTools while logged in as a test student, on a throttled mobile profile.
Test a sales page logged out, then the same course's first lesson logged in. The gap between the two numbers tells you whether your problem is front-end weight or server time.
Logged-in lesson pages are built fresh for every view, and LCP cannot start until the HTML arrives. If time to first byte is over about 800 milliseconds, fix that first: a current PHP version, enough PHP workers, and a persistent object cache. Our guide to object caching with Redis for LifterLMS covers the biggest single improvement for logged-in pages.
The Featured Video and Featured Audio fields on courses and memberships, and the Video Embed Url field on lessons, are turned into players with WordPress's wp_oembed_get(). Unlike embeds placed in post content, which WordPress caches in post meta, that function does not store its result, so a page that is not served from a page cache can spend time asking the video provider for embed code before it can finish. The plugin at the end of this article caches that response.
The player itself is often the LCP element on a lesson. Iframes from YouTube or Vimeo load a lot of script. Consider a lightweight facade, a poster image that loads the real player on click, for lessons where the video is not watched immediately, and use your video host's own performance options.
fetchpriority="high" to the image it believes is the LCP candidate; check in the HTML that your theme or page builder has not overridden that.INP is about main-thread work. On course sites the main sources are page builders, analytics and marketing tags, chat widgets, and scripts from plugins that load everywhere.
LifterLMS contributes some of that. Its front-end asset loader enqueues the main llms script, an AJAX helper, the checkout form script, a popover library, and three jQuery UI components (tooltip, datepicker and slider) on every front-end page, including pages with no course content. Logged-in users also get the notifications script. Each is small, but together with jQuery they add parsing and execution time on low-end phones. LifterLMS's own front-end code does not use those jQuery UI components, which makes them a safe first target, as long as your add-ons do not rely on them.
Beyond LifterLMS:
LifterLMS's own video wrapper reserves a 16:9 box for featured videos, so a featured video should not shift the page. The usual sources of layout shift on course sites are elsewhere:
LifterLMS's pop-up notifications for logged-in users are positioned over the page rather than inserted into it, so they are not a layout-shift source by themselves.
This plugin caches oEmbed responses for a day and removes the three jQuery UI scripts outside quizzes, checkout and the student dashboard. Install it on staging, test checkout, quizzes and any add-ons with front-end interfaces, then measure again.
<?php
/**
* Plugin Name: LLMSX Core Web Vitals Tweaks
*/
defined( 'ABSPATH' ) || exit;
// 1. Cache oEmbed HTML (featured videos, audio) for a day, so an uncached
// page view does not wait on the video provider.
function llmsx_oembed_cache_key( $url, $args ) {
return 'llmsx_oembed_' . md5( $url . wp_json_encode( $args ) );
}
add_filter( 'pre_oembed_result', function ( $pre, $url, $args ) {
if ( null !== $pre ) {
return $pre;
}
$cached = get_transient( llmsx_oembed_cache_key( $url, $args ) );
return ( false !== $cached ) ? $cached : null;
}, 20, 3 );
add_filter( 'oembed_result', function ( $html, $url, $args ) {
if ( $html ) {
set_transient( llmsx_oembed_cache_key( $url, $args ), $html, DAY_IN_SECONDS );
}
return $html;
}, 20, 3 );
// 2. LifterLMS enqueues three jQuery UI scripts on every front-end page.
// Core's own front-end code does not use them, so drop them where
// add-ons are unlikely to need them. Test your add-ons after enabling.
add_action( 'wp_enqueue_scripts', function () {
if ( is_singular( 'llms_quiz' ) || is_llms_checkout() || is_llms_account_page() ) {
return;
}
foreach ( array( 'jquery-ui-tooltip', 'jquery-ui-datepicker', 'jquery-ui-slider' ) as $handle ) {
wp_dequeue_script( $handle );
}
}, 20 );
The oEmbed cache uses transients, so with a persistent object cache it lives in memory, and without one it lives in the options table. The pre_oembed_result filter runs at priority 20 and respects any earlier result, which leaves WordPress's own handling of embeds from your site untouched.
Not much on its own. It adds a modest set of scripts and styles on every page. Themes, page builders, third-party tags and uncached server time usually matter far more.
Search Console reports on URLs it knows about, which are mostly public pages. Lessons behind a login matter less for rankings but just as much for learners, so measure them with DevTools.
Lightweight themes with native LifterLMS support start ahead, but your page builder and plugins decide the final result. Measure your own pages rather than relying on demo scores.
Field data covers the previous 28 days, so expect a few weeks after a fix before the report changes.
Pick one sales page and one lesson, measure both on mobile, and fix the slowest metric on each. If server time is the problem, start with object caching; if it is JavaScript, start with third-party tags. If you would like a full Core Web Vitals audit with fixes that do not break checkout, quizzes or progress tracking, our LifterLMS optimization service does exactly that. Send us the URLs that are failing and we will tell you what is holding them back.
What a LifterLMS backup must include, how often to take it, and how to restore without losing orders or double-charging students: a data map of…
Read article
A practical LifterLMS maintenance routine: what to check weekly, monthly, quarterly and yearly, where each check lives in the LifterLMS Status…
Read article
What it takes to scale LifterLMS to tens of thousands of students: hosting for concurrency, the database tables that grow fastest, running cron and…
Read articleSkip the trial and error. Get specialist LifterLMS help and ship faster.