What One Second of Delay Actually Costs Your Business
Picture an AC-repair company in Tampa. They run Google Ads, keep a clean Google Business Profile, and pull about 3,000 visitors a month to their site. Their booking page converts at 2%. That’s 60 leads a month, and at $400 revenue per booked call, they’re generating $24,000 in monthly revenue from web traffic.
Now add one second to the time that booking page takes to load.
According to the math, they don’t lose 60 leads. They lose roughly 4. But those 4 leads — gone silently, month after month — add up to $1,680 every month. Over a year, that’s $20,160 that never showed up in the revenue line. No invoice. No cancellation. Just a visitor who left before they ever called.
This is the cost of website speed, and most business owners have never calculated it for their own site.
Where the 7% Figure Comes From
In late 2024, Yottaa analyzed more than 500 million visits across 1,300-plus ecommerce sites — a dataset they published in their 2025 Web Performance Index. Their finding: each additional second of page load time costs roughly 3% of mobile conversions. Across a full second at the observed baselines, the cumulative conversion impact tracks consistently with the widely cited 7% per-second figure that the industry has used as a working benchmark.
That figure also has a historical anchor. A 2020 Deloitte study commissioned by Google — the “Milliseconds Make Millions” report — analyzed 30 million mobile user sessions across 37 brands in retail, travel, and lead generation. Deloitte found that a 0.1-second improvement in mobile load time correlated with an 8.4% increase in retail conversions and a 10.1% increase for travel sites. That study is five years old, but it was the first large-scale attempt to quantify the relationship precisely, and it remains foundational for that reason. The Yottaa 2025 data corroborates its direction with current figures.
One thing worth being honest about: Yottaa’s 500-million-visit dataset is observational. Sites with faster load times also tend to have better-built pages overall. The data shows an association, not a proven cause-and-effect chain. But when an association holds across 500 million visits from 1,300 sites, the pattern is robust enough to act on.
The Math Applied to Your Business
The worked example above is deliberately small — the kind of traffic a Tampa Bay HVAC company, dental practice, or law firm might actually see. The math scales to any size, but the principle is the same:
Baseline: 3,000 monthly visitors × 2% conversion rate = 60 leads × $400 = $24,000/month
One slow second: 60 leads × 7% = 4.2 leads lost × $400 = $1,680/month missing
Over 12 months: $20,160 left on the table — from a single second.
You can run this on your own numbers. Take your monthly visitors, your conversion rate, your average revenue per conversion. Apply 7%. That dollar figure is not the worst-case scenario. It is the middle-of-the-road estimate from a half-billion-visit dataset.
What changes if you have 10,000 monthly visitors instead of 3,000, or if your average lead value is $2,000 instead of $400? The multiplier stays the same. The dollar figure becomes impossible to ignore.
”My Site Feels Fast to Me”
This is the single most common reason business owners dismiss a performance problem that their customers are experiencing every day.
When you visit your own website from your office, you are almost certainly using a high-speed internet connection — fiber or cable — and the site loads with your browser’s local cache already warm. Your customers are typically on mobile, often on a 4G connection, cold cache, first visit. They experience a site that is measurably slower than the one you see.
Google’s own real-user measurement infrastructure — called CrUX, or Chrome User Experience Report — collects page-load data from actual Chrome users across the real conditions they visit sites in. That data frequently diverges from what a developer or owner sees in their own browser. If you want to know what your visitors actually experience, CrUX field data is the measurement that counts. You can see it at no cost in Google’s PageSpeed Insights tool, under “Field Data” — not the lab score that appears at the top.
A Real-World Reference: Nuvemshop’s 2026 Results
In a case study published by Google’s web development team, the Brazilian ecommerce platform Nuvemshop improved LCP — Largest Contentful Paint, the metric that measures how quickly the main visible content of a page loads — across their merchant stores. The share of stores with good LCP went from 57% to 96%. Measuring the same cohort of stores year-over-year, Nuvemshop recorded an 8.9% increase in mobile conversion rate and an 8.4% increase in cart engagement.
This was not a controlled A/B experiment. It was a before-and-after cohort study, and other variables changed alongside performance. The case study is clear about that. But the methodology — same stores, same users, year-over-year — is more rigorous than most before/after claims, and the outcome was published by Google’s Search Advocate John Mueller.
For a service business in Tampa Bay, the lesson is not “fix LCP and expect exactly 8.9% more conversions.” The lesson is that when a platform systematically improved this metric for nearly a thousand stores, conversion rates moved consistently in the same direction.
What LCP, INP, and CLS Actually Mean
These acronyms appear in every speed audit, and most business owners have never had them explained in plain language.
LCP (Largest Contentful Paint): How long it takes for the biggest visible piece of your page — usually a photo or a heading — to load. Think of it as “when does the page look done?” Google’s threshold is under 2.5 seconds at the 75th percentile of real users.
INP (Interaction to Next Paint): How fast the page visually responds after someone clicks or taps something. A slow INP means a visitor clicks “Book Now” and nothing happens for a beat — they assume the site is broken. Google’s threshold is under 200 milliseconds. (A full article on INP’s conversion impact is available in this series.)
CLS (Cumulative Layout Shift): Whether the page jumps around while loading — buttons that move, text that reshuffles, forms that shift. A high CLS score makes users tap the wrong thing and leave frustrated. Google’s threshold is under 0.1.
All three must pass simultaneously for your site to be considered “Good” by Google’s current standards. According to the HTTP Archive Web Almanac 2024 — which analyzed real-user data from 16.9 million websites collected in June 2024 — only 43% of mobile sites currently pass all three metrics. That means 57% of your competitors are failing at least one of these thresholds, and so might you.
What It Takes to Fix This
Website speed is not a redesign. It is a technical optimization project: identifying what is causing the slowdown (often third-party scripts, uncompressed images, or server response time), fixing those specific things, and verifying the results in real-user field data.
The timeline is weeks, not months. The tools exist to measure precisely what is failing and by how much. The calculation of what a fix might be worth — run on your actual traffic and revenue numbers — is the first step any honest assessment should produce.
The reason most business owners haven’t done this is not that it’s too expensive. It’s that no one has ever handed them a dollar figure attached to their specific site. Once you have that number, the decision is not complicated.
The Free Assessment
At iServU, the engagement starts with a technical assessment of your site against current CrUX field data — not a lab score, not an estimate from your own browser. We identify what’s failing, what it’s costing in estimated lost conversions, and what the fastest path to fixing it looks like.
No account manager. No handoffs. Gabriel Lopez-Seco, the founder, runs the assessment directly and gives you the numbers without a sales deck attached.
If the math above made you want to run it on your own site, that’s exactly what the assessment does.
Book a free assessment at iServU →
Sources
- Yottaa (2025). 2025 Yottaa Web Performance Index. yottaa.com
- Deloitte / Google (2020). Milliseconds Make Millions. meshwithus.com.au
- Google web.dev / Nuvemshop (2026). Core Web Vitals Case Study: Nuvemshop. techwyse.com
- HTTP Archive / Web Almanac (2024). Web Almanac 2024: Performance. httparchive.org
