SEO & Search Rankings

I treat SEO as an engineering problem, not a marketing trick. I trained in mathematics, so I would rather measure than guess — “the data says” beats “it might be.” When a site is not ranking, there are usually clear, measurable reasons behind it. The job is to find those reasons, prioritise them correctly, and fix them in the right order.

That is the difference between SEO as opinion and SEO as engineering. Opinion says, “Maybe we need more blog posts.” Engineering asks: is Google crawling the right pages, are those pages indexed, are they fast enough, do they satisfy the intent, are they internally linked, do they deserve to rank compared with the current winners, and can users actually convert once they arrive?

Most SEO specialists give advice. I can also do the implementation — because I am a developer too. If a technical fix is needed I do not say “pass this to your developer”; I make the fix. If the infrastructure is the problem, I can rebuild a fast, search-friendly site from scratch. This page is an example of that: clean HTML that ships almost no JavaScript and loads instantly.

That matters because SEO problems often live inside the site itself: templates, rendering, redirects, canonicals, JavaScript, internal links, content structure, product filters, image handling, schema, and performance. If the person diagnosing the issue cannot touch the system, the work slows down and the recommendations often stay theoretical. I prefer closing the loop: diagnose, implement, measure, iterate.

Why SEO is engineering, not guesswork

Rankings are the result of hundreds of signals — but none of them are magic. Each is measurable: is a page crawled, is it indexed, how fast does it load, does it match search intent, does it cover the topic more completely than competitors, do other sites trust it? I answer those with data and start with the highest-impact issues. Pareto logic: most of the gains come from a small share of the work — you have to find that share first.

For example, a site might have hundreds of pages, but the ranking ceiling may come from one basic issue: Google is not reaching the important pages efficiently. Or the content might be good, but the templates produce duplicate titles and confusing canonical tags. Or the site might be visually impressive but so JavaScript-heavy that key content appears late, loads poorly, or is harder for search engines to process.

Good SEO is not about doing every possible thing. It is about identifying which constraints matter most right now. Sometimes that is technical. Sometimes it is content depth. Sometimes it is internal linking. Sometimes competitors have built authority that the site has not earned yet. The work changes depending on the bottleneck, but the principle stays the same: measure first, then act.

This is also why I do not like vague SEO theatre. “Fresh content,” “more keywords,” or “optimise meta tags” can be useful, but only if they address the actual limiting factor. If the page is not indexed, rewriting a heading will not solve the problem. If the site is slow because of render-blocking scripts, adding more copy will not fix Core Web Vitals. If the content does not match search intent, backlinks alone will not make users happy.

The process: how I build rankings

1. Technical audit and foundation

Everything starts with a crawl. Crawlability, indexing status, Core Web Vitals, rendering method, duplicate content, broken canonical/hreflang, orphan pages and internal linking. The goal is to lift the ceilings that silently cap your rankings. Details: technical SEO.

This first stage is about making sure the site can actually compete. I look at what search engines can discover, what they can understand, and what they are choosing to index. A site can look healthy on the surface while hiding serious problems: important pages buried too deep, duplicate versions competing with each other, incorrect canonical tags, hreflang conflicts, thin pages indexed at scale, or useful pages that are not linked from anywhere important.

The foundation matters because every later SEO activity depends on it. Content, links, and conversion work all perform better when the technical base is clean. If the crawl path is broken, internal authority is wasted. If pages are slow, users leave earlier. If the rendering model hides the main content, Google may not see the page the way users do. The audit is not a PDF for its own sake; it is a roadmap for removing friction.

2. Keyword and intent mapping

We choose keywords by intent and conversion, not just volume. “Looking for information” and “ready to buy” searches go to different pages. I map the keyword gap — terms competitors rank for but you do not — and start with winnable long-tail.

Volume alone can be misleading. A broad keyword may bring traffic, but if the visitor is at the wrong stage, it may not produce leads or revenue. A lower-volume long-tail query can be far more valuable when it reveals clear intent. For example, someone searching a general definition needs education; someone comparing services, requesting pricing, or looking for a specialist is much closer to action. Those users should not land on the same generic page.

Intent mapping also prevents pages from fighting each other. Each important query cluster needs a clear destination. Informational searches may need guides, explainers, or comparison pages. Commercial searches may need service pages, category pages, product pages, or local landing pages. The goal is to build a structure where every page has a job and every keyword group has a natural home.

3. On-page and content

Title, H1, heading hierarchy, internal linking, structured data and, above all, content that serves the topic better than competitors. Using synonyms and related entities correctly lets a single strong page rank for an entire cluster of keywords.

This is where precision matters. A page should make its topic obvious to both users and search engines. The title should frame the search intent. The H1 should confirm the subject. The headings should organise the argument logically. Internal links should show how the page fits into the wider site. Structured data should clarify the page type and relevant entities.

But on-page SEO is not just formatting. The real question is whether the page deserves to be one of the best results for the query. Does it answer the obvious follow-up questions? Does it explain the process? Does it show expertise? Does it compare options honestly? Does it make the next step clear? A thin page with a few keywords inserted is not enough. A strong page covers the subject naturally and completely.

4. Authority off-page

White-hat, sustainable authority: digital PR, genuine mentions, industry relationships. I stay away from bought link schemes — a short-term spike means a long-term penalty.

Authority still matters, but it has to be built in a way that survives. Search engines are good at recognising patterns that exist only to manipulate rankings. Cheap link schemes may look attractive because they promise speed, but they introduce risk and usually do not build a real brand. I would rather earn signals that make sense: relevant mentions, useful resources, partnerships, expert contributions, and visibility in places your audience already trusts.

Off-page SEO should support the business, not just a link graph. A genuine mention in the right industry context can help rankings, referral traffic, trust, and brand recognition at the same time. That is the kind of authority worth building.

5. Measurement and iteration

Real data from Search Console and GA4: crawling, impressions, clicks, conversions. I measure what works and shift effort there. Reporting is transparent — you see what you are paying for and what you are getting.

SEO is not a one-time checklist. Search results change, competitors move, pages age, and new opportunities appear. Measurement tells us what is improving and what needs adjustment. Search Console shows how Google is discovering and displaying the site: impressions, clicks, query growth, page performance, indexing signals. GA4 shows what users do after they arrive: engagement, conversion paths, and business outcomes.

Transparent reporting matters because SEO should not feel like a black box. You should know what was changed, why it was changed, and what the data says afterwards. If something works, we do more of it. If something does not move the needle, we stop wasting time and redirect effort.

Technical SEO: the layer most agencies skip

This is where my development background pays off. Most agencies stop at meta tags. I go deeper:

• Core Web Vitals — LCP (largest contentful paint), CLS (layout shift) and INP (interaction delay). I measure these with real-user data (CrUX) and diagnose the cause (render-blocking scripts, unoptimized images, heavy JavaScript).
• Crawl budget and indexing — log analysis shows whether Google’s bot is spending its time on your money pages or wasting it on parameter URLs.
• Rendering — whether content is generated on the server (SSR/SSG) or in the browser with JavaScript decides whether Google ever sees it. This is the most commonly missed issue on modern sites.
• Structured data (schema.org) — tells search engines exactly what a page is; essential for rich results and AI citations.

All of this gets its own page: technical SEO.

The reason this layer gets skipped is simple: it requires technical fluency. It is easier to change a title tag than to diagnose a rendering pipeline, inspect server responses, reduce JavaScript, fix hydration problems, analyse logs, restructure internal links, or rebuild templates. But those are often the issues that create the biggest ranking ceiling.

Core Web Vitals are a good example. Saying “make the site faster” is not a diagnosis. LCP might be poor because the hero image is oversized, because fonts block rendering, because the server responds slowly, or because a client-side app delays the main content. CLS might come from images without dimensions, injected banners, late-loading fonts, or unstable layout components. INP might be caused by heavy JavaScript, third-party scripts, or inefficient event handlers. Each cause needs a different fix.

Rendering is another common blind spot. Modern sites often rely heavily on JavaScript frameworks. That can be fine when implemented well, but dangerous when important content, links, or metadata are only available after client-side rendering. SSR/SSG can make content immediately available in the HTML, while browser-generated content can introduce delays and uncertainty. If Google does not reliably see the content, the page cannot reliably rank for it.

Site speed is the foundation of SEO — and I build the site

A fast site ranks better and converts better. I build sites with modern tools like Astro and Next.js, shipping near-zero JavaScript where possible — the result is pages that load instantly with perfect Core Web Vitals. Most SEO specialists say “make your site faster” and move on; I can make it faster or rebuild it. Having SEO and development in one hand removes the “let’s tell the developer” delay.

Speed is not cosmetic. It affects crawling, indexing, user experience, conversion, and perceived trust. A slow site creates friction at every step. Users hesitate, forms feel heavier, product pages feel less responsive, and mobile visitors suffer most. Search engines also have to process the site efficiently, and bloated pages make that harder.

The practical approach is to avoid unnecessary weight in the first place. Serve clean HTML. Optimise images. Load only the JavaScript that is truly needed. Use static generation where appropriate. Keep templates simple. Avoid turning every page into a heavy application when the user just needs fast information and a clear path to action.

This is where development and SEO belong together. If the best SEO recommendation is “rebuild the page template,” I can do that. If the site needs a lighter framework, better image handling, cleaner internal linking, or server-rendered content, I can implement it directly instead of producing a recommendation that waits in someone else’s queue.

Content and topical authority

Google ranks a topic, not a keyword. To become an authority in a field, we build a cluster: the topic’s main page, the hub, linked to the long-tail sub-pages, the spokes, around it. This page is a hub; technical SEO and e-commerce SEO are its spokes. Each serves a clear intent, and they all pass authority to one another.

A hub-and-spoke structure works because it mirrors how expertise is understood. One broad service page can explain the main subject, but it cannot go deeply into every specialist subtopic without becoming unfocused. Spoke pages allow depth: technical SEO can cover crawlability, indexing, rendering, schema, and performance in more detail; e-commerce SEO can focus on category architecture, product pages, faceted navigation, and commercial search behaviour.

Internal linking connects the cluster. The hub introduces the topic and points users to deeper resources. The spokes answer specific intents and link back to the hub. Over time, this creates a clearer topical footprint. Search engines can see not just one isolated page, but a structured body of content around the subject.

Topical authority also protects against shallow content. Instead of publishing disconnected articles, we build a system. Each page has a reason to exist, a target intent, and a place in the architecture.

GEO / AEO: visibility in AI search

Search is no longer just ten blue links. Answer engines like AI Overviews, ChatGPT and Gemini answer questions directly and cite their sources. To appear in them you need clear entity definition, solid structured data and citable, evidence-based content. GEO (Generative Engine Optimization) is a space most competitors have not taken seriously in 2026 — as a developer, this is exactly where we can get ahead.

The same principles that make a page strong for search also help it become useful for answer engines: clarity, structure, evidence, and authority. A page should define what it is about, who it is for, what problem it solves, and what facts support the answer. Ambiguous marketing copy is less useful than precise, citable information.

Structured data helps machines understand the page. Clear headings help extract answers. Strong internal linking helps define the entity and its relationship to other topics. Evidence-based content gives answer engines something worth citing. This is not about gaming AI systems; it is about making your expertise understandable to both humans and machines.

Because I work at the technical layer as well as the content layer, GEO / AEO is not treated as a separate gimmick. It is built into the architecture: clean HTML, structured data, fast pages, clear entities, and content that answers real questions.

Which businesses?

The approach changes with the type of business. For online stores, category, product and faceted structure are critical — e-commerce SEO. For corporate sites, information architecture and the technical foundation; for local businesses, Google Business Profile and local signals. Each has its own playbook.

An e-commerce site often needs careful control of crawl paths, filters, duplicate pages, indexable category pages, product schema, internal linking, and commercial keyword mapping. A corporate site usually needs stronger service architecture, clearer positioning, better technical performance, and pages that match high-value B2B search intent. A local business needs consistency, proximity signals, reviews, local landing pages where appropriate, and a well-optimised Google Business Profile.

The point is that SEO is not one universal checklist. The same principles apply, but the priorities change. A local service provider does not need the same structure as a large online store. A SaaS company does not need the same content plan as a restaurant. The right playbook depends on how your customers search and how your site needs to support that journey.

Why work with me

• A mathematician’s lens — decisions follow the data, not a hunch.
• Strategy and execution — I do not just advise; I make the technical fixes and, if needed, build the site.
• SEO + Google Ads in one hand — we plan organic and paid search together for full search coverage.
• Transparency — real data, clear reporting, no overpromising. I will not guarantee first place; I deliver measurable progress.

The main difference is ownership. I am not interested in vague recommendations that sound good but never get implemented. If a problem is technical, I can solve it technically. If the content structure is weak, I can rebuild it logically. If paid search and organic search are competing or missing opportunities, I can plan them together so the whole search presence works as one system.

I also prefer honest expectations. SEO is competitive, and nobody serious can guarantee first place. What can be guaranteed is a disciplined process: diagnose the real issues, prioritise the highest-impact work, implement cleanly, measure results, and keep improving.

Engagement model and pricing

Three core models: a one-off technical audit + roadmap, ongoing consulting/management, or project-based work, for example a new site plus SEO setup. A free initial analysis tells us where to start, then I give you a clear quote. No blind packages — you know upfront what you are paying for and what to expect.

A one-off technical audit is useful when you need a clear diagnosis and a practical roadmap. Ongoing consulting or management is better when SEO needs continuous implementation, measurement, and iteration. Project-based work fits when the site itself needs to be rebuilt, migrated, restructured, or launched with SEO built in from the start.

I do not believe in selling the same package to every business. Some sites need a deep technical cleanup. Others need better content architecture. Others need authority building, local optimisation, e-commerce restructuring, or a faster website. The first step is understanding where the real constraint is, then defining the work around that.

If you want to talk about why your site is not where it deserves to be, get in touch. I usually reply within a day.