Page speed is the most frequently measured and least frequently understood metric in web development. Everyone knows fast is good. Almost nobody can explain what fast actually does to their bottom line.
We can, because we measure it.
The Numbers That Matter
Google has published the data repeatedly. Here is what the research consistently shows:
- A 1-second delay in mobile page load reduces conversions by up to 20% (Google/SOASTA, 2017 — confirmed in subsequent studies)
- 53% of mobile visits are abandoned if a page takes longer than 3 seconds to load (Google DoubleClick)
- A 100ms improvement in load time increased conversion rates by 8.4% for retail sites and 10.1% for travel sites (Akamai, 2017)
- Core Web Vitals — Google’s page experience metrics — are a confirmed ranking factor. Faster sites rank higher.
These are not edge cases from laboratory conditions. These are aggregate measurements across millions of real users on real connections. The relationship between speed and revenue is not debatable. It is measured, published, and reproduced.
What 100ms Feels Like
A hundred milliseconds is below the threshold of conscious perception. You cannot feel 100ms. Your brain can — it registers the difference as “this feels responsive” versus “something is slightly off.”
Jakob Nielsen’s research on response times, published in 1993 and still cited because nothing has changed, identifies three thresholds:
- 100ms — the user feels the system is reacting instantly. No perception of delay.
- 1 second — the user notices a delay but their flow of thought is uninterrupted.
- 10 seconds — the limit of keeping the user’s attention. Beyond this, they leave.
Most business websites live in the 2-5 second range. They are past the flow-of-thought threshold and approaching the attention limit. Every visitor is making a subconscious calculation: is this worth waiting for?
When your page loads in under a second, that calculation never happens. The visitor is already reading your content.
Where the Time Goes
A typical WordPress site with a theme and 10-15 plugins loads like this:
- DNS lookup — 50-100ms
- TLS handshake — 50-100ms
- Server processing (PHP execution, database queries, plugin hooks) — 200-800ms
- HTML transfer — 50-200ms (depending on page size)
- CSS/JS download and parsing — 200-500ms (a typical theme ships 500KB-2MB of assets)
- Render blocking — the browser cannot paint the page until critical CSS and JS are parsed
- Third-party scripts — Google Fonts, analytics, chat widgets, social embeds — each adding 100-300ms
Total: 2-5 seconds. Sometimes longer. And this is on a good connection. On a slow or congested mobile connection — common in rural areas, crowded tourist spots, and developing markets — double or triple those numbers.
Now here is how our sites load:
- DNS + TLS — same 100-200ms (unavoidable network physics)
- HTML transfer — under 50ms. Our pages are small static files, not server-rendered PHP.
- CSS + JS parsing — under 100ms. Total assets: ~30KB. The browser parses this almost instantly.
- Content fetch from API — 100-300ms (parallel with rendering, so it does not block the initial paint)
Total: under 500ms on a good connection. Under 1 second on a weak connection. The page is interactive before a traditional WordPress site has finished its server processing step.
The Architecture That Makes This Possible
We achieve these numbers through three architectural decisions:
1. No server-side rendering. Traditional WordPress generates HTML on every request — executing PHP, querying the database, running plugin hooks, rendering templates. Our SPA is a static file. Nginx serves it from disk in microseconds. There is no processing step.
2. Minimal assets. A React site starts at 150KB of JavaScript before a single line of application code. A WordPress theme ships 500KB-2MB of CSS and JS. Our entire application — HTML, CSS, and JavaScript — is under 30KB. That is not a typo. Thirty kilobytes.
3. Parallel content loading. The SPA shell renders instantly (the navigation, the layout, the skeleton). Content from the WordPress API loads in the background. The user sees a complete, interactive page while the content streams in. There is no blank white screen waiting for the server to finish thinking.
None of these techniques are novel. They are well-understood web performance patterns. The difference is that we build every site this way from the start, rather than trying to optimise a heavy site after the fact.
The Business Case — Worked Example
Consider a dive shop website receiving 10,000 visitors per month. Industry average conversion rate for tourism websites: 2-4%. Let us use 3%.
At 3% conversion on 10,000 visitors: 300 bookings per month.
Now apply Google’s research. A site loading in 5 seconds versus 1 second loses approximately 20% of potential conversions — not through poor design or weak copy, but through impatience.
20% of 300 bookings = 60 lost bookings per month.
If the average booking value is $50 (a reasonable number for a discover scuba course or fun dive), that is $3,000 per month in lost revenue — $36,000 per year — because the page took 4 seconds longer to load.
This is conservative. The real numbers for a busy dive operation in the Philippines are often higher. And this calculation only accounts for the initial page load. Every subsequent page navigation on a fast SPA is instant (no server round-trip), which compounds the engagement advantage.
What PageSpeed Insights Measures (And Why It Matters)
Google’s PageSpeed Insights measures six key metrics. Here is what they mean in plain language:
- First Contentful Paint (FCP) — how quickly the user sees something on screen. Our target: under 1 second.
- Largest Contentful Paint (LCP) — how quickly the main content (hero image, headline) is visible. Our target: under 1.5 seconds.
- Total Blocking Time (TBT) — how long JavaScript blocks the page from responding to user input. Our target: under 50ms (because we ship almost no JavaScript).
- Cumulative Layout Shift (CLS) — how much the page jumps around while loading. Our target: zero. Static layouts do not shift.
- Speed Index — how quickly the visible content fills the viewport. Our target: under 1.5 seconds.
- Time to Interactive (TTI) — when the page is fully usable. Our target: under 1 second.
Our sites consistently score 90+ on PageSpeed Insights. Not because we run optimisation tools after building — because the architecture is inherently fast. You cannot optimise a heavy page down to 30KB. You have to start with 30KB.
The SEO Compound Effect
Page speed affects more than user experience. Google confirmed Core Web Vitals as a ranking signal in 2021. A faster site, all else being equal, ranks higher than a slower one.
Higher rankings mean more traffic. More traffic at the same conversion rate means more customers. More customers at the same speed advantage means the gap between you and your slower competitors widens over time.
This is not a one-time benefit. It compounds. Every month your site is faster, the traffic advantage accumulates. After a year, the difference between a 90-score site and a 50-score competitor is not just a ranking position — it is a fundamentally different volume of organic traffic.
What This Means For Your Business
If your website takes more than 2 seconds to load on mobile, you are losing customers. Not because your design is bad or your content is weak — because the visitor’s patience ran out before your server finished thinking.
The fix is not a caching plugin, a CDN bolt-on, or a speed optimisation service layered on top of a heavy site. Those are band-aids. The fix is an architecture that is fast by default — one where speed is not optimised but inherent.
That is what we build. Every site. Every time.
Key Facts: Page Speed and Business Performance
- 100ms improvement in load time = 8.4% increase in retail conversion (Akamai)
- 53% of mobile users abandon sites loading over 3 seconds (Google DoubleClick)
- Core Web Vitals confirmed as Google ranking factor since 2021
- Typical WordPress site: 500KB-2MB assets, 3-5 second load time
- DESIGN-R SPA: ~30KB assets, sub-1-second load time. 10-60x smaller, 3-5x faster.
- PageSpeed score: DESIGN-R sites score 90+. Industry average for WordPress: 50-70.
Frequently Asked Questions
Q: Can’t I just add a caching plugin to make my WordPress site faster?
A: Caching plugins help — they can reduce server processing time significantly. But they cannot reduce the weight of your theme and plugin assets. A cached WordPress page still sends 500KB+ of CSS and JavaScript to the browser. Our entire site is 30KB. Caching narrows the gap; architecture eliminates it.
Q: Does page speed really affect my Google ranking?
A: Yes. Google confirmed Core Web Vitals as a ranking signal in 2021. Sites that meet all three thresholds (LCP, CLS, INP) receive a ranking boost. The effect is modest for any individual signal but compounds over time, especially in competitive niches where other ranking factors are similar.
Q: How much faster can my site realistically get?
A: If your current site loads in 3-5 seconds, a headless architecture with minimal JavaScript can bring that to under 1 second — a 3-5x improvement. The exact improvement depends on your current stack, but we have never built a site that scored below 90 on PageSpeed Insights.
Q: Is a fast site enough to improve conversions?
A: Speed alone does not guarantee conversions — you still need clear messaging, strong calls to action, and a product people want. But speed removes a barrier. It ensures that visitors who would have converted are not lost to impatience before they see your offer.
Q: What about images? Don’t large images slow down any site?
A: Images are the largest assets on most pages, regardless of architecture. We handle this with lazy loading (images load only when they enter the viewport), responsive sizing (serving smaller images on mobile), and modern formats (WebP where supported). The key difference is that our page is interactive before images finish loading — the visitor can read and navigate while images stream in.
Originally published at DESIGN-R Intelligence
Top comments (0)