SEO · · 9 min read
How Technical SEO Became the Most Underrated Growth Channel
Most marketing budgets go on content and advertising. Far less goes on whether a search engine can fetch, render and understand the pages that content lives on, and whether a visitor who clicks will wait for them. That infrastructure layer is technical SEO. It is unglamorous, it is measurable, and — on the evidence below — it is where a Malaysian business with a modest budget still finds its cheapest gains in search.
What the question is
A client usually asks us one of two things: "is our site slow enough to be hurting us?" and "is any of this still worth doing now that Google answers questions itself?" Both are empirical questions with public data behind them — Google’s own documentation, the HTTP Archive’s annual Web Almanac, a Google-commissioned Deloitte study on speed and revenue, and two large click-through studies on AI Overviews. We work through that data here and then say what we would do with it.
A word on scope: almost all of this evidence is global. No Malaysia-specific Core Web Vitals or click-through dataset exists that we would trust, so the global figures are the baseline here.
What the data says about speed
Google’s Core Web Vitals are three field metrics: Largest Contentful Paint for loading, Interaction to Next Paint for responsiveness and Cumulative Layout Shift for visual stability. Google’s web.dev documentation sets the "good" bar at 2.5 seconds, 200 milliseconds and 0.1 respectively, measured at the 75th percentile of real page loads and split by mobile and desktop.
The 2024 Web Almanac, which draws on Chrome’s real-user dataset, found only 43% of origins with good Core Web Vitals on mobile against 54% on desktop. The gap is responsiveness and loading: 74% of mobile origins passed INP against 97% on desktop, and 59% passed LCP against 74%; layout stability was the one metric where mobile did better (79% against 72%). The median mobile page carried 1,209 milliseconds of Total Blocking Time where Google’s guidance is under 200 — six times over budget on main-thread work.
| Metric | Good | Needs improvement | Poor |
|---|---|---|---|
| Largest Contentful Paint | ≤ 2.5 s | 2.5–4.0 s | > 4.0 s |
| Interaction to Next Paint | ≤ 200 ms | 200–500 ms | > 500 ms |
| Cumulative Layout Shift | ≤ 0.1 | 0.1–0.25 | > 0.25 |
Does any of this move revenue? The best public evidence is still the Google-commissioned Deloitte study "Milliseconds make millions", which tracked mobile site data from retail, travel, luxury and lead-generation brands in Europe and the US over four weeks. A 0.1 second improvement in mobile speed was associated with an 8.4% rise in retail conversions and a 9.2% rise in retail average order value, a 10.1% rise in travel conversions, 8.6% more pages per session for luxury brands and an 8.3% improvement in bounce rate on lead-generation pages. The study does not say how many brands took part and is observational, so we treat the direction as reliable and the exact percentages as indicative.
What the data says about structure
Structured data — schema.org markup, ideally as JSON-LD — is how a page tells a search engine what it is rather than leaving it to infer. Google’s introduction to structured data cites four companies’ own measurements: Rotten Tomatoes added markup to 100,000 pages and saw a 25% higher click-through rate on them; Food Network converted 80% of its pages and saw visits rise 35%; Nestlé measured an 82% higher click-through rate on pages shown as rich results; Rakuten saw users spend 1.5 times longer on marked-up pages. These are self-reported by large publishers and quoted by Google, so they are the optimistic end of the range — but the direction is consistent and the mechanism is plain: a result showing ratings, prices or steps occupies more of the screen than a blue link.
Adoption is still uneven. The 2024 Web Almanac’s structured data chapter found JSON-LD on 41% of pages, up from 34% two years earlier, with Microdata flat at around 26%. The common JSON-LD types are generic — WebSite on 12.7% of mobile pages, Organization on 7.2%, BreadcrumbList on 5.7% — while the types that earn rich results are rare: Product on 0.8%, FAQPage, VideoObject and Event on about 0.3% each. Most sites with schema have the kind that changes nothing on the results page. Google describes structured data as a way of helping it understand content and qualify for rich results, not as a ranking factor, and we do not sell it as one.
Why this is a search problem and not just a user problem
Google is unusually plain about how much page experience counts. Its documentation says Core Web Vitals "are used by our ranking systems", that there is "no single signal", that it "always seeks to show the most relevant content, even if the page experience is sub-par", and that "trying to get a perfect score just for SEO reasons may not be the best use of your time". The honest reading is that speed is a tie-breaker among relevant pages — which, for a Malaysian SME competing with a dozen similar firms for the same commercial queries, is exactly its situation.
Rendering is the second mechanism. Google’s JavaScript SEO documentation describes three phases — crawl, render, index — and says a page "may stay on this queue for a few seconds, but it can take longer than that" before headless Chromium runs its JavaScript. A 2024 study by Vercel and MERJ instrumented more than 100,000 Googlebot fetches, mostly on nextjs.org, and found every HTML page with a 200 status fully rendered, with a median crawl-to-render delay of 10 seconds and a 75th percentile of 26 seconds — but a 90th percentile of about three hours and a 99th of about 18 hours, and no rendering at all for 3xx, 4xx and 5xx responses. The old fear that Google cannot see client-rendered content is unsupported; the real costs are a tail of hours before content is indexed, and the fact that other crawlers often do not run JavaScript — which is why Google still recommends server-side or pre-rendering.
Crawl budget is the third and the most over-sold. Google’s own guide says it matters for sites of a million or more pages that change weekly, or ten thousand or more that change daily — a "rough estimate", not a threshold. A Malaysian company site of a few hundred URLs has no crawl-budget problem; it has a duplicate-URL, soft-404 or slow-server problem, which the same guide lists as the things that waste the crawl you do get. The 2024 Web Almanac found 0.8% of pages with conflicting canonical tags, double the rate of 2022 — a small number, and exactly the plumbing error that sends a page’s ranking signals to the wrong URL.
The AI Overviews question
Here the news is bad for informational content, and the studies agree on direction if not magnitude. Ahrefs compared Search Console click-through rates for 300,000 keywords between December 2023 and December 2025: where an AI Overview now appears, the top organic result’s CTR fell from 0.073 to 0.016 — about 58% below where the trend for comparable informational keywords without an Overview would have put it. Seer Interactive, tracking 3,119 informational terms across 42 client organisations and 25.1 million organic impressions from June 2024 to September 2025, found organic CTR on queries with an Overview fell from 1.76% to 0.61% (a 61% drop) — but CTR on queries without one also fell 41%, from 2.72% to 1.62%. People are clicking less everywhere.
| Organic position | CTR change |
|---|---|
| First | −58.0% |
| Second | −50.8% |
| Third | −46.4% |
| Fifth | −32.6% |
| Eighth | −28.8% |
| Tenth | −19.4% |
Two details matter for strategy. Seer found that brands cited inside the AI Overview kept about 35% more organic clicks than those that were not (0.70% against 0.52%) and 91% more paid clicks — a correlation, the authors stress. And both studies describe the queries most affected as informational. Transactional and local intent — "custom software company Kuala Lumpur", "ERP for a manufacturing SME" — is where a Malaysian business earns its leads, and where AI Overviews appear least. Our reading: the top of the funnel is being absorbed into the results page, and clear, well-structured, fast pages are the price of being the source that gets cited rather than summarised and skipped.
A worked example
The point is the shape of the return, not the number. Speed work on an existing site is a few days of engineering with no ongoing cost, and the gain compounds across every channel that lands on the page — paid, social and email as well as organic.
How to start
We run the same audit on every site before recommending anything, and most of it needs no special tooling.
- Read the Core Web Vitals report in Search Console and the field data in PageSpeed Insights. Field data, not lab scores, is what Google uses; fix mobile first, because that is where the failures are.
- Find the LCP element on each key template. The Web Almanac found it is an image on 73% of mobile pages; keep that image in the initial HTML, never lazy-loaded (16% of mobile pages still do), with fetchpriority="high".
- Cut main-thread work: defer or remove third-party scripts, split bundles, and check INP on real phones rather than a developer’s laptop.
- Check indexation: one canonical per page, meaningful status codes (a 404 for a removed page, not a 200 that says "not found"), a current sitemap with lastmod, and no staging noindex left on live content.
- Add the structured data that changes the result for your page type — Product, LocalBusiness, FAQPage, Article, Event — validated in the Rich Results Test; skip the generic types.
- Render server-side or pre-render anything that must be indexed quickly or read by non-Google crawlers.
- Re-measure after each change and keep the numbers; a before-and-after in Search Console is the only proof the work paid.
Costs, risks and what we would do
The cost side is gentle: fixes to an existing site are bounded engineering tasks, and the platform matters more than the budget. The 2024 Web Almanac’s CMS chapter found only 40% of WordPress sites passed Core Web Vitals on mobile against 73% on Duda and 60% on Squarespace — not because WordPress is slow by nature but because an accumulation of plugins and themes usually is. A site that fails on its platform’s defaults may be cheaper to rebuild than to tune.
The risks are over-promising and over-building. Speed is a tie-breaker, not a ranking lever, and Google says so; structured data earns a richer result, not a higher one; chasing a perfect lab score past the "good" thresholds is spend without return. And every new tracking script, hero video or plugin re-opens the budget, so the vitals belong in the release checklist, not an annual audit.
What we would do, in order: pass the three vitals on mobile at the 75th percentile; fix indexation and canonicals; add the specific schema types that earn rich results for the commercially important page types; render the content server-side; and only then write more content, aimed at the transactional and local questions where a clear answer still earns the click.
Key takeaways
- Only 43% of mobile origins passed all three Core Web Vitals in 2024, so a site that does pass is already ahead of most of its competitors on Google’s tie-breaker.
- A 0.1 second mobile speed gain was associated with 8.4% more retail conversions in the Deloitte study — the cheapest conversion lift most sites have left.
- Structured data earns richer results, not higher rankings: Google’s cited case studies show 25–82% higher click-through, but only the specific types (Product, FAQ, Event) change anything, and those are on under 1% of pages.
- Google renders JavaScript reliably but with a tail of hours, and other crawlers often do not — server-side rendering is still the safe default for anything that must be found.
- AI Overviews cut the top result’s clicks by 58–61% on informational queries; the answer is to be the fast, well-structured source that gets cited, and to aim content at the transactional and local intent where the Overview rarely appears.
Sources
- Web Vitals
- Largest Contentful Paint (LCP)
- Interaction to Next Paint (INP)
- Cumulative Layout Shift (CLS)
- Web Almanac 2024 — Performance
- Web Almanac 2024 — SEO
- Understanding page experience in Google Search results
- Milliseconds make millions
- Introduction to structured data markup in Google Search
- Web Almanac 2024 — Structured Data
- Understand the JavaScript SEO basics
- How Google handles JavaScript throughout the indexing process
- Large site owner’s guide to managing your crawl budget
- Update: AI Overviews reduce clicks by 58%
- AIO impact on Google CTR: September 2025 update
- Web Almanac 2024 — CMS