Technical SEO Expert for Websites That Should Rank but Don’t

Search engines discover pages by crawling links, render them to see what a user would see, decide whether to index them, and only then rank them against user intent and search queries. Every step can fail silently. Pages blocked by robots.txt never get crawled. Pages with a wrong canonical tag get merged into another URL. Pages with rendering issues, where the main content is loaded by JavaScript that the crawler never executes, get indexed as empty shells. Pages that load slowly on mobile devices lose the user before the content appears, and page experience signals notice.

That is why technical SEO is the first thing to check when a website is not showing on Google, when organic traffic drops after a redesign, when a site migration loses rankings, or when hundreds of pages sit in the “crawled, currently not indexed” report. It is also why technical work compounds: fixing crawl errors, redirect chains and index bloat helps every page on the site at once, which is something no single blog post can do. On larger sites, crawl budget is a genuine constraint, and a technical SEO audit that removes low-value URLs from the crawl frees search engines to spend their time on the important pages. For a full audit of the whole site before any technical work starts, including content, links, local and AI visibility, see the SEO audit consultant page.

There is a commercial side too. Site speed affects conversion rates as much as it affects rankings; a slow product page loses sales whether or not it ranks. Structured data earns rich results that lift click-through rate. A clean URL structure and internal linking help both users and search engines find what they came for. And in 2026, AI search adds another layer: AI crawlers, AI Overviews and generative engines read the same HTML, follow the same links and rely on the same structured data as traditional search, so the technical foundation now decides whether a business appears in AI-generated answers as well as in the search engine results pages.

Everything starts with the crawl. A search engine crawler arrives at your website, reads robots.txt to learn what it is allowed to request, fetches the XML sitemap to learn what you think matters, and then follows internal links from page to page. If the site architecture buries a service page six clicks deep with no internal links pointing to it, the crawler may never arrive. If a robots.txt file blocks the CSS and JavaScript folder, the crawler fetches the HTML but cannot render it properly, so it never sees the layout, the images or the text that JavaScript was supposed to inject. If a meta robots tag says noindex on a template that was copied from the staging site, whole sections vanish from the index and nobody notices until the enquiries dry up.

Indexability is the second gate. Google can crawl a page and still decide not to index it. The index coverage report in Google Search Console lists the reasons: duplicate without user-selected canonical, alternate page with proper canonical tag, crawled but not indexed, discovered but not indexed, soft 404, redirect error, blocked by robots.txt, server error 5xx. Each label points to a different technical problem. “Crawled, currently not indexed” usually means thin or near-duplicate content. “Discovered, currently not indexed” on a large site usually means crawl budget: the crawler knows the URL exists but has not bothered to fetch it because it has thousands of low-value URLs to get through first.

The fix list from a crawl and index audit is specific. Remove parameter URLs, tag archives, internal search results and filtered category pages from the index with canonical tags or noindex, and stop linking to them where you can. Return a real 404 or 410 status code for pages that are gone, instead of a 200 that shows a “not found” message, because a soft 404 wastes crawl budget and confuses the index. Consolidate duplicate pages with 301 redirects. Rebuild the sitemap.xml so it contains only indexable, canonical, 200-status URLs, split into multiple sitemaps on big sites so you can see which section has coverage problems. Check the hreflang tags on international sites, because a broken hreflang set can make the wrong language version rank in the wrong country. Then read the server logs, because logs show what crawlers actually requested, how often, and which status codes they got back, which is the only honest picture of crawl behaviour.

Sites that serve more than one country add hreflang, geotargeting and duplicate-version checks to that list; the international SEO services page covers that layer in full.

Page speed is the technical SEO issue business owners can feel. A website that takes six seconds to load on a mobile phone loses visitors before the first headline appears, and search engines measure that experience through Core Web Vitals: Largest Contentful Paint (LCP), which is how long the biggest visible element takes to appear; Cumulative Layout Shift (CLS), which is how much the page jumps around while it loads; and Interaction to Next Paint (INP), which replaced First Input Delay (FID) and measures how quickly the page responds when someone taps or clicks. The targets are an LCP under 2.5 seconds, a CLS under 0.1 and an INP under 200 milliseconds, measured on real users’ devices through the Chrome User Experience Report rather than on a fast laptop in an office.

That distinction between field data and lab data matters. Lab tools simulate a single load on a throttled connection and produce a score, which is useful for finding problems. Field data is what your actual visitors experienced over the last twenty-eight days, and it is what the page experience signals use. A site can score 95 in the lab and still fail Core Web Vitals in the field because most of its traffic arrives on older Android devices over patchy 4G in a car park. Speed optimisation therefore starts with the mobile performance of the pages that actually get traffic, not with the homepage on a desktop.

The fixes are unglamorous and effective. Compress and resize images, serve them in next-generation formats like WebP or AVIF, and set explicit width and height attributes so the layout does not shift when they arrive. Lazy-load images below the fold and never lazy-load the hero image, because that delays LCP. Inline the critical CSS for the first screen and defer the rest. Defer or delay non-essential JavaScript, especially third-party scripts: chat widgets, heat-map tools, social embeds, tag managers and tracking pixels are the most common cause of a slow, jumpy page. Preload the fonts you actually use and let the browser fall back to a system font while they load. Enable server-side caching, GZIP or Brotli compression and HTTP/2 or HTTP/3, put a CDN in front of static assets, and check Time to First Byte (TTFB), because if the server takes 1.8 seconds to start responding, nothing on the front end can rescue the LCP.

Hosting is part of technical SEO whether the hosting company likes it or not. Shared hosting that struggles under load returns 5xx server errors to crawlers and users at peak times, and a crawler that keeps hitting errors slows down. Database queries that run on every page load, bloated themes, page builders that output ten nested divs for every button, and forty plugins each loading their own stylesheet all add up. The right order of work is measure, fix the biggest bottleneck, measure again, and keep a written record of what changed and what it did, because speed work is easy to undo by accident the next time someone installs a plugin.

Site architecture is how your pages relate to each other, and it decides where link equity flows. A flat architecture keeps the important pages, the ones that sell, within two or three clicks of the homepage, with clear category pages in between. A deep architecture buries them. Search engines treat the number of internal links pointing to a page as a signal of how much you value it, so a service page linked once from a footer menu will always struggle against a competitor’s page that is linked from the navigation, the homepage, three blog posts and a breadcrumb trail.

URL structure should be readable, consistent and permanent. Short, lowercase, hyphenated URLs with the keyword in them, one URL per page, no session IDs, no capital letters, no trailing-slash inconsistency, no www and non-www versions both returning 200. Every time a URL changes, a 301 redirect has to map the old address to the new one, and redirect chains, where A redirects to B which redirects to C, need to be collapsed to a single hop because each hop leaks link equity and slows the crawl. Broken internal links and links to redirected URLs are cleaned up at the source, in the HTML, not just patched with more redirects.

Internal linking is the cheapest ranking lever most sites never pull. Contextual links inside body copy, with descriptive anchor text that says what the target page is about, pass relevance as well as authority. Orphan pages, pages with no internal links at all, are invisible to crawlers that discover content by following links, and they appear on more sites than anyone expects: old landing pages, campaign pages, pages that were removed from a menu during a redesign. Breadcrumb navigation, marked up with BreadcrumbList schema, gives every page a path back to its category and the homepage and shows up in the search results as a clean trail instead of a raw URL. Pagination on category and blog archives needs self-referencing canonicals and crawlable links, not infinite scroll that the crawler cannot follow.

Duplicate content is usually an architecture problem rather than a plagiarism problem. The same product reachable at three URLs through different category paths, a printer-friendly version, an HTTP version alongside HTTPS, filtered category pages that differ only by sort order, tag archives that reproduce whole posts: each creates multiple URLs competing for one query, which is keyword cannibalisation by accident. Canonical tags, consistent internal links and a sensible taxonomy resolve most of it. Where two pages genuinely target the same search intent, the answer is to merge them and redirect, not to leave both fighting.

Structured data is code, usually JSON-LD placed in the page head, that tells search engines exactly what a page is about in a vocabulary they already understand: this is an Organization with this name, logo, phone number and address; this is a Service offered in these areas; this is a Product with this price, availability and review rating; this is an FAQPage with these questions and answers; this is an Article by this author published on this date; this is a LocalBusiness open at these hours. Search engines use it to understand entities and relationships, and to decide whether a page qualifies for rich results: the star ratings, FAQ dropdowns, breadcrumbs, sitelinks, product prices and event details that make a listing take up more space and earn more clicks than a plain blue link.

The technical part is getting it right. Schema markup has to match what is visible on the page, or it is treated as spam. Nested types have to be connected properly, so the Service points at the Organization that provides it and the Organization has a single @id that every page references. Product markup needs offers, price, currency and availability in the correct fields. FAQ markup only earns the rich result on certain kinds of pages now, and review markup only counts when the reviews are genuinely on that page and genuinely about that thing. Every implementation gets validated with the Rich Results Test and the Schema Markup Validator, then checked again in Search Console’s enhancement reports a few weeks later, because a warning that appears after a template change is easy to miss.

Beyond rich results, structured data has become one of the clearest signals for AI search. Generative engines and AI Overviews assemble answers from pages they can parse cleanly, and explicit entity markup, meaning who you are, what you do, where you do it and what it costs, reduces the guesswork. On this site every service page carries Organization, Service, FAQPage and BreadcrumbList schema for exactly that reason, and the same approach is applied to client sites whatever the platform.

A website migration is the single most common way a business loses organic search traffic overnight, and it is almost always avoidable. Site migrations come in several flavours: a new design on the same platform, a move to a new content management system, a change of domain, a switch from HTTP to HTTPS, a merger of two sites, or a restructure of the URL structures across an entire ecommerce catalogue. Each one changes the signals search engines have already learned, and each one needs a migration checklist that starts weeks before launch rather than a panic the morning after.

The work begins with a comprehensive audit of the existing site: a full crawl exported to a spreadsheet, the pages that currently earn organic traffic and backlinks pulled from web analytics and Search Console, the keywords each page ranks for, and a benchmark of rankings, indexed pages and Core Web Vitals so that any change after launch can be measured against a baseline. Then comes the redirect mapping, old URL to new URL, one to one wherever possible, with 301 redirects rather than 302s, no chains, and no lazy catch-all redirect that sends every old page to the homepage. Every page with external links pointing at it gets special attention, because those backlinks are the authority the new site inherits.

On the staging site, before anything goes live, the new templates are crawled and checked for the classics: a noindex tag left on from development, a robots.txt that still blocks everything, canonical tags pointing at the staging domain, missing meta tags and title tags, images without alt text, internal links that still point at old URLs, structured data that broke when the template changed, and load speed that regressed because the new theme ships with more JavaScript than the old one. Launch day is monitored hour by hour: server logs, crawl errors, index coverage, redirect status codes and the first rankings movement. Recovery, where it is needed, is a matter of finding which signals were lost and restoring them, which is far easier when the pre-migration benchmark exists.

Modern websites are built on a wide range of technologies, and each brings its own technical SEO issues. WordPress has its own set of default behaviours to correct, from tag archives to attachment pages, covered in full on the WordPress SEO services page. Sites built on JavaScript frameworks render their content in the browser, which means the crawler first sees an almost empty HTML file and has to execute the JavaScript to discover the text, the links and the images. Search engines can do this, but rendering is queued, expensive and imperfect: content that appears only after a click, a scroll or an API call may never be indexed, and client-side routing can leave every page sharing one title tag and one URL in the crawler’s view. Server-side rendering, static generation or pre-rendering the important pages, plus a check of what the rendered HTML actually contains, resolves most of it. The rule is simple: if the primary content is not in the HTML that the server returns, assume it is invisible until proven otherwise.

Ecommerce SEO multiplies every technical problem by the size of the catalogue. Faceted navigation generates thousands of filtered URLs for colour, size, price and brand that all show near-identical product lists. Product variants spawn duplicate pages. Out-of-stock products get deleted and return 404s, throwing away the backlinks and rankings they had earned, when a 200 page with a “notify me” option or a 301 to the parent category would keep the value. Category pages need crawlable pagination, unique copy above the product grid and canonical tags that agree with the internal links. Product schema with price, availability and reviews turns plain listings into rich results. Site speed on product pages is directly tied to revenue, and an online store that loads slowly on a phone is paying for traffic it cannot convert. E-commerce technical SEO is therefore an ongoing process of controlling what gets crawled, consolidating what gets indexed and making the pages that matter fast.

International SEO adds hreflang, the set of tags that tells search engines which language and country version of a page to show to which searcher. Broken hreflang, where the tags are missing return links, point at redirected URLs or use the wrong region codes, is one of the most common findings on multi-country sites, and the result is the wrong version ranking in the wrong market. Country targeting, ccTLDs versus subfolders versus subdomains, currency and language handling, and duplicate content across near-identical English versions for the UK, the US and Australia all need deliberate decisions rather than defaults.

Mobile is no longer a special case; it is the case. Mobile-first indexing means search engines evaluate the mobile version of a page for ranking, so content hidden on mobile, menus that only exist on desktop, and images that only load at desktop widths are effectively not there. Mobile usability covers tap targets, viewport settings, font sizes, intrusive interstitials and the layout shifts that happen when banners and cookie bars push content around. Responsive design solves the structural side; performance work solves the speed side; and a mobile-friendliness check on the pages that earn revenue, not just the homepage, catches what a desktop-only review misses.

AI-driven search has changed what a search result is, but it has not changed how content gets found. AI Overviews, AI Mode and the answer engines assemble responses from pages that were crawled, rendered and understood in exactly the way traditional search requires. That makes technical SEO the entry ticket to AI search visibility. A page that is slow, blocked, rendered only by JavaScript or stripped of structured data is as invisible to a generative engine as it is to a classic crawler, and often more so, because AI crawlers tend to have less patience with rendering than the main search crawler does. The AI SEO agency page covers the rest of that work: entity clarity, answer-ready content and the mentions that get a business cited.

Generative engine optimisation, sometimes shortened to GEO, and answer engine optimisation therefore start with the same audit: clean, semantic HTML with one clear H1 and a logical heading structure; answer-shaped sections where a definition, a price or a process is stated plainly in the first sentence; structured data that names the organisation, the service, the area served and the offer; fast pages that return complete content in the initial HTML; and crawler access rules that allow the AI systems you want to appear in. Some sites publish an llms.txt file to describe their content to AI systems, alongside robots.txt; it costs nothing and does no harm, though the evidence for its effect is still thin. Entity clarity matters more than any single tag: the same business name, address, phone number and service descriptions everywhere, on the site and across the wider web, so that AI systems and knowledge graphs resolve the brand to one entity with confidence.

What changes for measurement is the reporting. Impressions in AI features do not behave like classic rankings, and the click-through rate from an AI answer is lower than from a top organic result. The honest picture combines Search Console data, web analytics, referral traffic from AI platforms and a regular manual check of which queries produce an AI answer and whether the site is cited in it. The wider picture of how AI Overviews are changing clicks is covered in the guide to AI Overviews and SEO, and the organic side of the service sits under SEO services.

A technical SEO audit is a health check of the whole website against the things search engines actually measure, turned into a prioritised list of fixes with a clear owner for each one. The first step is access: partner access to Google Search Console and Google Analytics 4 in your name, never your passwords, plus read access to the hosting or content management system where the fixes will be made. Then the site is crawled in full, the same way a search engine crawler would see it, and the crawl data is compared against the sitemap, the analytics data and the server logs to find the gaps: pages that exist but are never crawled, pages that are crawled but never indexed, pages that are indexed but never visited, and pages that are visited but never convert.

The audit itself covers crawlability and indexability, site speed and Core Web Vitals, website architecture and internal linking, on-page elements like title tags, meta descriptions and heading structure, structured data, mobile usability, security and HTTPS, duplicate content, redirects and broken links, JavaScript rendering, international targeting and, increasingly, AI search readiness. Every finding is checked by hand rather than copied from an automated tool, because auditing tools flag hundreds of “issues” that do not matter and miss the handful that do. A missing alt attribute on a decorative icon is not the same as a canonical tag pointing every product page at the homepage, and a report that lists them side by side with the same severity is worse than useless.

What comes back is a written document in plain English: the problems ordered by impact, what each one is costing in traffic or conversions, how to fix it, how long it takes, and whether it is something you can fix yourself for nothing. Where I do the implementation, the changes are made in accounts and systems you own, tested on staging where one exists, and confirmed live. Where your developer does it, the tickets are written so that a developer who has never heard of SEO can complete them correctly. Either way, a weekly written report shows what changed, what moved in Search Console and analytics, and what is next.

Where a development team needs ongoing SEO direction rather than a one-off fix list, the SEO consulting services page describes how tickets, briefs and reviews are delivered month to month.

Prices are published, fixed and agreed in writing before any work starts. Technical SEO can be bought on its own as a one-off project or as part of a monthly retainer where it runs alongside content, local SEO and link building. Ad spend, where there is any, is always paid by you directly to the platform; nothing here is a percentage of a media budget.

International clients pay in USD at $800, $1,500 and $5,000 a month for the equivalent plans. Everything is month to month with no long-term contract, and everything built, from the audit document to the schema and the redirect map, stays yours if you stop. The full price list and the fixed-price packages are on the pricing page and the packages page; if you want to understand what a sensible budget looks like before talking to anyone, the UK SEO cost guide sets out what the market charges and why.

Technical SEO is a specialism inside a specialism, and plenty of digital marketing agencies sell it without anyone on the team who can read a server log. The questions below separate a technical SEO specialist from a marketing manager reading from a checklist, whether you are hiring a freelance technical SEO consultant, a technical SEO agency or an in-house SEO analyst.

And the warning signs, learned the hard way by businesses that came here after a bad experience:

BoltClicks was founded in 2026 and is run by one specialist, so there is no wall of client logos here and no case studies with names attached until clients agree in writing to be named. What there is instead is this website, which is built the way I build client sites, and you are welcome to test it: crawl it, run the Core Web Vitals, read the structured data on any page, check the index coverage against the sitemap, and look at how the service, location and guide pages link to each other. The process page explains how an engagement runs from the first message to the weekly report, and the honesty page sets out what is and is not promised.

The work is remote, from Lahore, for UK and US businesses of any size, from a single-location trade business with a twelve-page site to an online store with a large catalogue. Remote is the point rather than a compromise: nothing about crawling a site, fixing a canonical tag or mapping a migration requires being in the same building, and it keeps the fee going into the work. Communication is through WhatsApp, email and a weekly written report, and the answer to any message arrives within one working day.

Technical SEO is the foundation, not the whole house. Once search engines can crawl and index the site quickly and understand what each page is about, the rest of an SEO strategy starts to work: keyword research that maps one search intent to one page, on-page SEO that writes title tags, meta descriptions and headings people actually click, content that answers the questions your target audience asks, local SEO and Google Business Profile work for businesses that serve an area, and link building and digital PR that build authority and brand visibility. Off-page SEO without a solid technical base sends hard-won backlinks into pages that cannot rank; technical SEO without content ranks pages that have nothing to say.

It also underpins everything else in a digital marketing strategy. Paid search campaigns land on the same pages, and landing page speed affects Quality Score and cost per click. Social media traffic and email campaigns arrive on mobile and bounce from slow pages. Conversion rate optimisation depends on pages that render correctly and load fast enough for the visitor to see the form. Web analytics only report accurately when tracking is installed cleanly and the site does not fire duplicate events. Lead generation, ecommerce revenue growth and return on investment all sit downstream of a site that works, which is why technical SEO comes first in the order of work here, whether the goal is organic search, paid media or both.

For businesses that want the whole programme, the SEO services page covers the full organic strategy, local SEO covers the map pack and Google Business Profile, SEO copywriting covers the content side, and Google Ads management covers paid search. For businesses that only need the technical foundation fixed and handed back, the one-off SEO Foundation package exists precisely for that.

After enough audits the same problems appear again and again, across industries, platforms and company sizes, from healthcare and education to retail, travel, fintech and SaaS. Most of them are invisible from the front end, which is why they survive redesigns and agency changes. If you want a quick self-check before talking to anyone, these are the ones to look for first.

Technical SEO comes with its own language, and most of the jargon hides simple ideas. The short definitions below cover the terms that appear in audits, in Search Console and in the conversations with developers, so that the weekly report never needs a translator.

Crawling and indexing

Crawling is a search engine’s bot, Googlebot for Google, fetching your pages by following links. Crawlability is whether it can. Indexing is the search engine storing a page in its database so it can appear in results; indexability is whether a page is allowed to be stored. Indexation is the state of being indexed. Crawl budget is how many URLs a crawler will fetch on your site in a given period; on small sites it rarely matters, on large ones it decides which pages get seen. Crawl errors and crawling errors are failed fetches, usually 404s or server errors. Robots.txt is the text file at the root of a domain that tells crawlers which paths they may request. Meta robots and the X-Robots-Tag header do the same job per page, including noindex and nofollow. A sitemap, usually sitemap.xml, is a list of the URLs you want indexed, with lastmod dates. Log file analysis reads the server’s record of every request, including every crawler visit. Index coverage is the Search Console report showing which URLs are indexed and why the rest are not. Rendering is the crawler executing JavaScript and CSS to see the finished page rather than the raw HTML.

URLs, redirects and duplicates

A canonical tag, written as rel canonical, tells search engines which URL is the master copy when several show the same content; a self-referencing canonical points at the page itself. A 301 redirect is a permanent move, a 302 is temporary, and a redirect chain is several hops in a row. A status code is the server’s answer to a request: 200 for OK, 301 and 302 for moved, 404 for not found, 410 for gone, 500 and 503 for server errors. A soft 404 is a “not found” page that returns 200. Duplicate content is the same or near-identical content reachable at more than one URL. Keyword cannibalisation is two of your own pages competing for the same query. URL structure is how addresses are organised; URL parameters are the bits after a question mark. Hreflang tags map language and country versions of a page to each other. HTTPS is the encrypted version of HTTP, signalled by a valid SSL certificate; mixed content is an HTTPS page loading HTTP resources.

Speed and experience

Core Web Vitals are three user-experience metrics: Largest Contentful Paint for loading, Cumulative Layout Shift for visual stability and Interaction to Next Paint for responsiveness. First Input Delay was the older responsiveness metric, and First Contentful Paint measures when anything first appears. Time to First Byte is how long the server takes to begin responding. Field data comes from real users; lab data from a simulated test. Render-blocking resources are scripts and stylesheets the browser must load before it can paint the page. Critical CSS is the small amount of styling needed for the first screen. Lazy loading defers images and iframes until they are about to be seen. Minification strips whitespace from code; compression shrinks files in transit; caching stores copies so repeat visits are faster; a CDN serves files from a location near the visitor. Mobile-first indexing means the mobile version is the one evaluated for ranking, and mobile usability covers viewport, tap targets and text size. Bounce rate and user engagement are analytics measures of whether visitors stay.

Structure and markup

Site architecture or website architecture is how pages are organised and linked; click depth is how many clicks a page sits from the homepage. Internal linking is links between your own pages; anchor text is the clickable words; link equity is the ranking value a link passes; an orphan page has no internal links pointing to it. Breadcrumbs show the path to a page. Pagination splits long lists across pages. Faceted navigation is the filter system on ecommerce category pages. Structured data, written as schema markup from the schema.org vocabulary, usually in JSON-LD, describes entities on a page; rich results or rich snippets are the enhanced listings it can earn. Title tags, meta descriptions and heading tags (H1, H2, H3) are the HTML tags that describe a page to search engines and searchers; alt text describes an image; metadata is the collective term. Semantic HTML uses elements for their meaning. JavaScript SEO is the discipline of making script-rendered content indexable, and server-side rendering is the usual fix.

Measurement and the wider field

Google Search Console, once called webmaster tools, is Google’s free report on how it crawls, indexes and ranks your site; its URL inspection tool shows how a single page was fetched. Google Analytics measures what visitors do. A rank tracker records positions for chosen keywords over time. SERPs are search engine results pages; organic search is the unpaid results; organic traffic is the visits they produce; impressions are how often a listing was shown and click-through rate how often it was clicked. Ranking factors are the signals an algorithm uses; algorithm updates are the changes to them. Domain authority and similar scores are third-party estimates of link strength, not Google metrics. Backlinks are links from other sites; backlink analysis reviews them; a disavow file tells Google to ignore harmful ones. On-page SEO is everything on the page, off-page SEO is everything elsewhere, and technical SEO is the layer that makes both possible. Local SEO targets searches with local intent and the map pack. Topical authority is the depth of coverage a site has on a subject. Generative engine optimisation and answer engine optimisation are the emerging terms for earning citations in AI-generated answers, and AI visibility is the measure of how often that happens.