Lighthouse and Core Web Vitals Comparison

Google’s Lighthouse and Core Web Vitals are two different methods for measuring your website's performance and user experience. There are several differences in how Lighthouse and Core Web Vitals measure website performance, what metrics are used, what constitutes a passing score, and which tools you can use to run each report. This article compares the two approaches so you can make an informed decision about when to use each one.

How Performance Scores are Determined

Lighthouse

Lighthouse measures the performance, accessibility, SEO, and other best practice quality indicators of your web pages. To do this, Lighthouse uses lab data that is generated in a fixed testing environment. This lab data is used to simulate visits to your web page with a fixed CPU speed, 3G internet connection, and a Moto G4 mobile device. The use of a slightly underpowered CPU, slower internet connection, and less powerful smartphone reflects the real world conditions of users visiting your site.

Core Web Vitals

Core Web Vitals are a set of three metrics that attempt to measure and summarize the overall user experience of visiting your web pages by focusing on loading speed, interactivity, and visual stability. These metrics use field data, which is real-world data anonymously sent to Google from actual users when they visit a specific page.

We recommend that you rely on Core Web Vitals data as much as possible because real-world aggregated data is a better indicator of performance than point-in-time data based on a controlled set of lab conditions. And, the score you receive on Core Web Vital's metrics can make a difference in how prominently you appear within Google’s own search engine results pages for your targeted search queries.

Metrics and Scores

Both Lighthouse and Core Web Vitals use different metrics to generate their reports. Each set of metrics have optimal thresholds that dictate whether or not a page is considered passing.

Lighthouse

Lighthouse uses six different metrics to generate its report:

Largest Contentful Paint (LCP). Measures your page’s loading time and checks how long it takes for the largest above the fold element (image, text, and so on) to load. The optimal threshold is less than 2.5 seconds.

Cumulative Layout Shift (CLS). Measures the visual stability of a page load by focusing on unexpected layout shifts not caused by a user interaction. Sometimes when a page loads, elements shift around unexpectedly and frustrate users. For example, you might load a page and start reading a paragraph, only for an image to load that pushes the paragraph down the page. The optimal threshold is less than 0.1 layout shift.

Total Blocking Time (TBT). Total amount of time that a page is blocked from responding to user input. These inputs include mouse clicks or keyboard strokes. The optimal threshold is less than 0.2 seconds.

First Contentful Paint (FCP). The time elapsed until the first piece of content loads on the page. The content must come from the page’s DOM (Document Object Model). The DOM includes standard page content like images and text. The optimal threshold is less than 1.8 seconds.

Speed Index. Measures the entire loading process for the visual parts of a web page by capturing a video of the page loading and checking the difference between frames. The total duration measures how long it takes to go from a blank screen to a complete page. The optimal threshold is less than 3.4 seconds.

Time to Interactive (TTI). How long it takes the page to become fully interactive so that it reliably responds to user inputs. The optimal threshold is less than 0.1 seconds.

The Lighthouse performance report converts raw scores into standard 0-100 values then uses a weighted average to calculate your website’s total performance score. A passing score for any individual metric or the total performance score is anything above 90.

Core Web Vitals

Core Web Vitals measures CLS and LCP in the same way as Lighthouse and uses the same optimal thresholds. In addition, Core Web Vitals also measures First Input Delay (FID), which measures the time it takes from the user's first interaction, such as clicking a button, until the browser responds to that action. The optimal threshold is less than 0.1 seconds. FID is similar to Lighthouse’s TBT metric and has the same optimal threshold. For a page on your website to be considered passing it needs to achieve the optimal threshold for each metric at the 75th percentile.

For more information, see Introduction to Core Web Vitals.