Developers love Lighthouse scores. Business owners rightly ask, "so what?" The honest answer is that the number itself is meaningless — but what it measures shows up in revenue.
The data, briefly
Google, Amazon, Walmart, Deloitte and Akamai have all published the same finding in slightly different ways: every additional second of load time reduces conversions by roughly 5–7% on mobile. It's one of the most reproducible findings in the entire discipline.
Some published numbers:
- Amazon reported that 100ms of latency cost them 1% in sales.
- Walmart found 2% conversion improvement for every 1-second speedup, up to 4 seconds.
- Deloitte found retail sites converting up to 8% higher after 100ms improvements.
What Lighthouse actually measures
The Performance score is a weighted average of five things:
- First Contentful Paint — when the browser paints anything.
- Largest Contentful Paint (LCP) — when the main content is visible. This is the one users notice.
- Total Blocking Time — how long the page is unresponsive to taps and clicks.
- Cumulative Layout Shift (CLS) — how much the page jumps around while loading. If a button moves right as you're about to tap it, that's CLS.
- Speed Index — a composite of how quickly content is visually complete.
Two of these — LCP and CLS — are also Core Web Vitals, which Google uses as a ranking signal. So a fast site converts better AND ranks better. Compounding.
Where the wins usually are
Most business sites lose the Performance score in the same three places:
Images. Almost always the biggest lever. Hero images uploaded at 4000px wide when they display at 1200. PNGs used instead of WebP or AVIF. No lazy loading. Fixing this alone often takes a site from 60 to 85.
Fonts. Loading five weights when the site uses two. Not preconnecting to the font host. Not using font-display: swap. Every unloaded font blocks the first paint.
Third-party scripts. A single unnecessary chat widget, analytics tag, or "social share" script can eat a 20-point performance score. Every third-party script is a bill the user pays in load time.
Why "98" instead of "100"
Chasing a perfect 100 is diminishing returns. The user experience at 98 is indistinguishable from 100, and the last two points often cost more engineering hours than the first 60. Aim for consistent 95+ across the metrics, and use those hours to build something users need instead.
The takeaway
Ask your developer for a Lighthouse report on your live site right now. If mobile Performance is below 80, you're leaving money on the table — probably 10–20% of your potential conversions. Fixing it is usually a two-day project, not a two-month one. The ROI is instant and it never turns off.