Examples
Illustrative Speed Optimization Examples
Pagelatix is a new studio, so there is no client work to show yet. Instead, here are three fictional sites built to show the kind of problems a speed audit finds and how the fixes are chosen. Every number on this page is a sample, not a result.
Oakline Goods: an online furniture store
Illustrative example, fictional store
The situation. A small furniture store on a hosted e-commerce platform. The home page opens with a five-slide carousel of full-width lifestyle photos. Product pages carry a review app, an upsell app, a size guide pop-up and three tracking pixels. Search Console reports product pages as poor on mobile for LCP and INP.
What an audit would look for. Which element is the LCP on the home and product templates, and which LCP subpart is longest. Which app scripts run before the main content appears. Whether collection pages lazy-load images in the first row.
Likely findings. The first carousel slide is the LCP element, but it is injected by a script, so the browser discovers it late. Product gallery images are uploaded at 3000 pixels and served at full size to phones. The review and upsell apps load on every page and run long tasks during load.
Fixes we would propose. Replace the carousel with a single hero image in the HTML, marked with high fetch priority. Serve responsive gallery sizes in a modern format. Load the review app when its section scrolls into view and the upsell app only on product pages. Keep cart and checkout untouched by any caching.
Harbor & Pine Advisors: a professional services site
Illustrative example, fictional firm
The situation. A five-person advisory firm on WordPress with a popular page builder theme. The site has 18 plugins, four font families with several weights each, a testimonial slider and a chat widget. Lab tests show a long element render delay and noticeable layout shift as fonts load.
What an audit would look for. Which stylesheets and scripts block rendering, how much of the page builder's CSS each page actually uses, and what each plugin loads on the front end.
Likely findings. Several plugins load styles on every page though their features appear on one. The fonts are fetched from a third party with no fallback metrics, so text reflows when they arrive. The chat widget loads immediately and competes with the main content.
Fixes we would propose. Reduce fonts to two families in WOFF2, self-hosted and preloaded, with adjusted fallbacks. Load plugin assets only where used. Remove the slider in favor of a static quote. Delay the chat widget until the visitor scrolls. Set up page caching appropriate to the host.
Field Notes Journal: an ad-supported content site
Illustrative example, fictional site
The situation. A long-form publishing site with display ads, a newsletter pop-up, embedded videos and a large tag manager container. Search Console shows article pages as poor for CLS on desktop and INP on mobile.
What an audit would look for. Which elements shift during load, whether ad slots and embeds reserve space, how many tags fire before the content is visible, and which interactions are slow on mobile.
Likely findings. Ad slots collapse and expand as ads arrive, pushing the article down. Video embeds have no set aspect ratio. The tag manager holds old campaign tags that still fire on every page, and a custom HTML tag runs a heavy script on load.
Fixes we would propose. Reserve space for each ad slot with a minimum height. Give embeds an aspect ratio and load them on demand. Pause and then remove unused tags, move non-essential tags to fire after load, and replace the custom HTML tag with a lighter template.
Why these are samples, not results
Scores like the ones above are lab results, and lab results vary between runs and between tools. Google's documentation notes that much of the variability in a Lighthouse score "is not due to Lighthouse" [1]. More importantly, a lab score is not what Google uses to judge Core Web Vitals; it uses real-user field data at the 75th percentile [2]. Real projects are reported with field data where it exists, and with dated lab measurements where it does not.
Sources
Opened and checked on November 3, 2026.
- Lighthouse performance scoring, Chrome for Developers. Primary documentation, last updated September 19, 2019.
- Web Vitals, web.dev (Google Chrome team). Primary documentation, last updated October 31, 2024.
Ready for a closer look?
Let’s Look at Your Website’s Speed
Send us the address and what feels slow. You get a short written first read by email, with no obligation and no promised score.
- No obligation
- Findings in writing
- Plain-English notes