Technical SEO Audit: Finding What Is Actually Holding a Site Back
An SEO audit is a diagnosis, run in the order search engines process a site — can a page be reached, is it indexed, is it understood, is it measured honestly. Here is every layer we check, what the deliverable is, and when an audit is the wrong thing to buy.
Searcora offers this as a paid SEO audit service: every layer described below, delivered as a prioritised list of what to fix first, not a 200-page export nobody reads. Quoted per site after a short conversation. If the audit finds your site is not the constraint, we will say so.
What an SEO Audit Actually Is
A diagnosis, run in the order search engines process a site.
An audit — some people call it a website audit, and it is the same work — finds out why a site is not earning the visibility its owner expects, in the order search works: a page must be reachable before it is indexed, indexed before it can rank, understood before it ranks for the right thing, and measured honestly before anyone can say a change helped.
That sequence matters more than the checklist. Rewriting the <title> tag on a page carrying a noindex directive changes nothing; building links to a page that canonicals to the homepage passes the benefit elsewhere. Most sites carry a stack of small contradictions rather than one dramatic fault, and the value is in sorting that stack into order.
It is deliberately the smallest sensible first step: finite, cheaper than an ongoing engagement, and yours to keep whether or not you continue with us.
Crawlability and Indexability
If a page cannot be fetched, or has been told not to be indexed, nothing else on it counts.
This pass establishes what a crawler reaches and what it has been told to do with it. The most expensive problems hide here, because they are invisible in a browser.
- robots.txt. Disallow rules inherited from a template, and blocked script or stylesheet folders that stop the page rendering as a visitor sees it. It governs crawling, not indexing.
- Meta robots and X-Robots-Tag. Stray page-level noindex, and the site-wide one that shipped from staging: a rebuild sits behind a blanket noindex, the launch checklist misses the line, and pages quietly leave the index.
- Canonicals. Tags pointing at the wrong URL, at a redirect, or — from a misconfigured theme — every page canonicalising to the homepage.
- Redirect chains and loops. Hops accumulated across migrations, redirects landing on a 404, loops nothing escapes. Each old URL should point once at its replacement.
- Orphan pages. URLs that exist, sometimes in the sitemap, that nothing on the site links to.
- Pagination and parameters. Endless filtered variants, sort parameters duplicating pages, paginated series leaving deeper items unreachable.
- Sitemaps and status codes. Sitemaps listing redirects, 404s and noindexed URLs; soft 404s returning 200 on an empty page.
Site Structure and Internal Linking
One page per intent, reachable in a few clicks, described by the links pointing at it.
Structure is how a site says which page answers which question, and internal links are the mechanism.
One page per intent. Each distinct thing customers search for needs one page that is unambiguously the best answer. Two half-pages are worse than one complete page.
Hub and spoke. A hub covers a subject in full and links down to the pages beneath it, each linking back up: local SEO supports narrower pages like Google Business Profile optimisation and the local SEO checklist. Those links make the group read as a subject rather than a pile of pages.
Click depth. Pages buried five levels down are treated as peripheral, by visitors too. A page nothing links to except the sitemap has been ranked last by its own site.
Anchor text. “Click here” describes nothing; descriptive anchors do. The opposite failure is worth naming too: identical exact-match anchors site-wide read as manipulation.
On-Page Content and Cannibalisation
What each page claims to be about — and whether two pages claim the same thing.
- Title tags. Missing, duplicated across a template, truncated, or padded with the business name and a string of towns.
- Meta descriptions. Not a ranking input, but the copy that argues for the click.
- Heading order. One <h1> stating the subject, then <h2> and <h3> reflecting the argument, not picked for font size. Skipped levels break the outline readers and machines scan by.
- Thin pages. Pages that exist to hold a keyword and answer nothing — a service page that names the service, asks for a call, and stops.
- Duplicate and near-duplicate content. Boilerplate repeated across templates, supplier descriptions used verbatim, location pages made by swapping the town name in otherwise identical text.
- Cannibalisation. Two pages chasing one query split links and relevance so neither is decisive. The check is which URL currently returns, and whether it keeps swapping. The fix is a decision — consolidate and redirect the loser, or genuinely separate them by intent.
Technical Performance and Core Web Vitals
Three published thresholds, and the build decisions that break each one.
Google publishes Core Web Vitals as user-experience metrics with defined “good” thresholds, assessed on real visits rather than a lab test over a fast connection. They are one input among many — a fast page about the wrong subject still does not rank.
| Metric | What it measures | Published “good” threshold | What typically breaks it |
|---|---|---|---|
| LCP — Largest Contentful Paint | Time until the main content has rendered | 2.5 seconds or less | Uncompressed hero images and video, render-blocking CSS and JavaScript, slow server response |
| INP — Interaction to Next Paint | How quickly the page responds after a tap or click | 200 milliseconds or less | Heavy JavaScript holding the main thread, stacked third-party tags, chat widgets and consent banners |
| CLS — Cumulative Layout Shift | How much the layout moves while loading | 0.1 or less | Images and embeds with no dimensions reserved, banners injected above content, late fonts resizing text |
These are build decisions, not settings to switch afterwards. An animation-heavy theme, a page builder emitting far more markup than the design needs, or images uploaded at camera resolution each set a ceiling no plugin removes. Where the ceiling is the platform, that is a web design conversation, and the audit says so.
Structured Data
Valid markup and eligible markup are two different things.
Correctness comes first: valid syntax, required properties present, no placeholder values left by the template that generated it. A frequent finding is two conflicting blocks on one page, from a theme and a plugin that do not know about each other.
Markup must describe what is on the page. Review markup for reviews shown nowhere on it, FAQ markup for questions visitors never see, offer markup on a page selling nothing. Describing content that is not visible breaches Google’s guidelines and can attract a manual action — reported as a violation, never as a tactic.
Valid is not eligible. Passing a validation tool means the markup is well formed and the feature is possible. It does not mean a rich result will be shown — eligibility is decided per query, and a flawless page can still show as an ordinary blue link. Any audit promising stars is selling something it does not control.
The Local Layer
For a business serving an area, this is usually where the largest gaps are.
For a local business the layers above are necessary but not sufficient, because a second system runs alongside the website. This layer checks the two agree.
- NAP consistency. Name, address and phone rendered identically on the site, in the markup, on the Business Profile and across the directories carrying the business.
- LocalBusiness schema. Present, accurate, the most specific applicable type, and matching what a visitor can read — not details existing only in code.
- Alignment with the Business Profile. Name, categories and services compared against the site; where they disagree, the disagreement is the finding. Profile work sits under Google Business Profile optimisation, and a suspended listing under Google Business Profile suspension.
- Location pages that are genuinely distinct. A page per area is legitimate when it says something true and specific about working there. A set differing only by a swapped place name is a doorway page, against Google’s guidelines, and a liability — see local SEO in London.
Three shortcuts are named as violations wherever they appear: keyword-stuffed business names, fake, virtual or unstaffed addresses, and review gating or incentivised reviews. Removing them joins the remediation list, with the honest note that undoing them can cost visibility first. The compliant way to earn and respond to reviews is covered under Google reviews for business.
Backlinks and Citations
A short list worth acting on, and a longer list worth ignoring.
| Worth reviewing | Mostly noise |
|---|---|
| Whether the pages that earn money have any links, or everything points at the homepage | Raw counts of backlinks and referring domains as a headline number |
| Links pointing at URLs that now 404 or redirect through a chain, reclaimable with one redirect or one email | Third-party “authority” and “spam” scores treated as though a search engine used them |
| Anchor patterns that look bought rather than earned, and paid links not marked as such | Toxic-link reports and routine disavowing, which the tool’s own documentation frames as a last resort |
| Citation accuracy on UK directories that matter: consistent details, duplicates merged, dead listings removed | Bulk submission to hundreds of low-quality directories, which mainly manufactures inconsistencies to clean up later |
The audit does not promise a number of links or offer to buy any. It identifies what is reclaimable, inaccurate or risky, and where genuine mentions come from: suppliers, trade bodies, local press, professional registers.
Analytics and Search Console Configuration
Checked last, and it can rewrite the conclusions of everything above it.
If the measurement is wrong, every earlier decision was made on bad information.
- Search Console properties. Whether one exists, whether it is domain-level or a prefix covering a single variant, and who still holds access.
- Analytics installation. Duplicate tags counting sessions twice, a tag missing from one template, filters that fail to exclude staff traffic.
- Conversion tracking. The common and costly case: a form or booking step changes, the confirmation URL changes with it, and the event recording a lead never fires again. Nothing breaks visibly. Reports keep arriving showing a flat line while the phone still rings, and months of decisions get made against a number that stopped being real. So it is verified by submitting the forms, not by reading a dashboard.
- What counts as a conversion. Sites where every page view is a goal, and sites where a phone call — the main way customers make contact — is recorded nowhere. Booking systems on other domains break the session and file customers as referral traffic.
Where tracking is broken, the audit says so before quoting any historical figure.
What You Actually Receive
A prioritised list with reasoning and an order of work — not a 200-page tool export.
A crawler export with every issue flagged red is not an audit. It is raw material, and handing one over transfers the judgement back to the client.
Every finding is written the same way: what was found, where it occurs, why it matters, what to do, and how to confirm it is fixed. They are ranked by size of problem against effort to fix, and ordered by dependency:
- Blockers. Anything stopping pages being crawled, indexed or served correctly, plus anything creating guideline risk.
- Measurement. Fixed early, because no later change can be evaluated without it.
- Structure and consolidation. Cannibalisation resolved, internal linking corrected, duplicates merged and redirected.
- Content and on-page. Titles, headings, thin pages, and the local layer.
- Performance and refinement. Core Web Vitals, structured data, and gains visible only once the larger problems are gone.
To be completely clear: an audit tells you what is wrong. It does not fix it. Implementation is separate work, and the document is written to be handed to whoever is best placed to do it — your developer, another agency, or us under a local SEO or web design engagement.
If this is the document you need: message WhatsApp +92 300 2462807 with your website address. You get a written scope and a quote before anything is charged — and if an audit is the wrong thing for you to buy, the next section says so.
When an Audit Is Not What You Need
Several situations where buying one would be the wrong call.
An audit is often the right first step, not always. We would rather say so at the enquiry stage than take the work.
- You already have one and have not implemented it. The constraint is capacity, not information — as it is if you know what is wrong and need hands rather than a diagnosis.
- The site is brand new. There is little to diagnose; building it correctly is the better use of the money — see web design.
- Your Business Profile is suspended. No website analysis reinstates a listing. Start at Google Business Profile suspension.
- The real problem is not search. If enquiries arrive unanswered, or nobody searches for the service where you work, an audit documents a healthy site attached to a problem it cannot solve.
- You are small, local and on a tight budget. Work through the free local SEO checklist first; if you finish it and are stuck on the technical layers, that is the moment for an audit.
- You want a guarantee. No consultant controls search rankings, and any promise of positions, traffic or map pack placement is a reason to walk away.
SEO Audit Services: What an Engagement Covers and How to Start
Scoped in writing, quoted per engagement, with access granted properly and never by password.
Searcora Digital Ltd is a UK registered company at 78 Dulverton Road, London, SE9 3RL, England, and the work is led by Muhammad Amir, Founder & Local SEO Consultant — Top Rated Plus on Upwork, a status Upwork awards and reviews rather than one that can be bought, with 600+ completed projects and a 99% Job Success score. An audit engagement covers:
- The eight layers above, across an agreed set of pages and templates rather than an undefined “whole site”.
- Manual review alongside crawl data: pages opened, forms submitted, journeys followed on a phone.
- A prioritised written document: each finding with its location, reasoning, fix and verification.
- An order of work in dependency sequence, and a walkthrough so the reasoning transfers, not just the list.
It does not cover implementation, content writing or reporting; those are scoped separately, and you are under no obligation to buy them from us.
What happens first is a short conversation: what you sell, where, what has been tried, and what a good outcome looks like in enquiries rather than positions. We then review what is publicly visible and return a written scope naming the pages and layers included, plus a quote. Nothing is charged before that scope is agreed. Pricing is quoted per engagement — a five-page site and a several-hundred-page site are not the same job, so no package prices are published.
What we need from you is access, granted the proper way:
- Permission to crawl the site, and a warning to whoever monitors server load.
- Search Console access, added as a named user on the property.
- Analytics read access, enough to inspect configuration and events.
- Business Profile manager access where the local layer is in scope, and CMS view access if templates need inspecting.
- Never passwords. All of these platforms invite named users with their own logins: safer for you, revocable in one click. If any agency asks you to email a password, treat that as a warning about how they work.
Where access cannot be granted we audit what is publicly visible and say which conclusions are unverified, rather than presenting a guess.
To start: message WhatsApp +92 300 2462807, email contact.amirjutt@gmail.com, or use the contact page with your website address and one line on what you expected search to be doing that it is not.