Important URLs return 200
Test: Open the homepage, priority services and contact journey without logging in. Check the final HTTP status after redirects.
If it fails: Repair accidental 404s and redirect old URLs to the closest useful replacement.
Printable audit framework
Website audit checklist: 30 checks that find real problems
Use this on a UK small-business website before a redesign, SEO campaign or paid-traffic launch. It checks whether useful pages can be found, believed and converted—not whether a tool can produce a long warning list.
Section 1
Crawling and indexing
Confirm that search engines can reach the right pages before judging keywords or copy.
Robots.txt is not blocking useful pages
Test: Read /robots.txt and test priority URLs in Search Console. A disallow rule controls crawling, not reliable removal from results.
If it fails: Remove broad disallow rules that hide pages or resources search engines need to render.
Indexable pages do not contain noindex
Test: Inspect the robots meta tag and X-Robots-Tag response header on every priority template.
If it fails: Use noindex deliberately for private, duplicate or low-value pages; remove it from pages intended to rank.
Canonicals name the preferred URL
Test: Every indexable page should normally use a self-referencing canonical. Check protocol, hostname, slash and parameter variants.
If it fails: Correct canonicals that point to redirects, errors or a different page without a genuine duplicate relationship.
The XML sitemap contains preferred URLs only
Test: Compare sitemap URLs with final status, canonical and robots state. Exclude redirects, errors and noindex pages.
If it fails: Generate the sitemap from the same source of truth used for page indexability.
Internal links expose priority pages
Test: Reach each commercial page through normal HTML links from the homepage, navigation, hub or relevant guide.
If it fails: Add descriptive links where a reader naturally needs the next page; do not rely on the sitemap alone.
Section 2
Search intent and page quality
Decide whether each URL earns a distinct place in the index and answers the query behind it.
Each page has one clear job
Test: Write the intended query, audience and next action beside every indexable URL. If two rows are identical, the pages may compete.
If it fails: Merge overlapping pages or give them materially different purposes and evidence.
The title is unique and descriptive
Test: Read titles without the website beside you. They should identify the subject and useful distinction without keyword repetition.
If it fails: Rewrite vague, duplicated or obsolete titles; keep the visible H1 aligned with the same promise.
The opening answers the search
Test: The first screen should confirm the service, problem, audience or answer before company history and generic claims.
If it fails: Move the direct answer and qualification details above supporting explanation.
Service pages explain the buying decision
Test: Look for scope, exclusions, process, price factors, location, evidence and what happens after contact.
If it fails: Add the missing decision information rather than padding the page to a word count.
Similar pages contain distinct evidence
Test: Compare templates side by side. Swapping only a town, trade or product name is not meaningful differentiation.
If it fails: Consolidate or add real local, service, project, policy or expert detail.
Claims can be substantiated
Test: Mark rankings, timeframes, percentages, client counts, awards, reviews and superlatives. Find the evidence for each.
If it fails: Remove or qualify anything that cannot be checked by a prospective customer.
Section 3
Mobile experience and performance
Test the journey on an ordinary phone and connection, not only a desktop design review.
The main action works without zooming
Test: Use the page at a narrow viewport. Check tap targets, sticky elements, menus, forms and keyboard behaviour.
If it fails: Make the primary call or enquiry action visible, labelled and easy to tap.
Core Web Vitals use field evidence where available
Test: Review Search Console or PageSpeed Insights field data for LCP, INP and CLS; use lab tests to diagnose, not to declare victory.
If it fails: Fix the largest real-user constraint and retest after deployment.
Images have dimensions and sensible weight
Test: Find oversized hero images, missing width and height, unsuitable formats and off-screen media loaded immediately.
If it fails: Resize at source, reserve layout space, compress and lazy-load only below-the-fold images.
The page remains usable without animation
Test: Enable reduced motion and test keyboard navigation. Confirm important content is not video-only or revealed solely by hover.
If it fails: Provide static alternatives and visible focus states.
Text and controls meet basic accessibility needs
Test: Check heading order, label associations, contrast, alternative text and error messages with an automated scan and manual keyboard pass.
If it fails: Fix barriers on templates first so one change improves every page.
Third-party scripts justify their cost
Test: List analytics, chat, advertising, video and widget scripts. Record who owns each one and what decision it supports.
If it fails: Remove unused tags and load non-essential services after the main content where practical.
Section 4
Trust and enquiry conversion
A ranking cannot create an enquiry if the visitor cannot verify the business or complete the next step.
A real person or business is identifiable
Test: Check the About page, footer, contact details and structured data for consistent, honest ownership information.
If it fails: Name who does the work and remove vague team, office or registration claims that are not evidenced.
Proof is attributable
Test: Reviews should have a real source; accreditations, memberships, case studies and statistics should be checkable.
If it fails: Do not publish placeholder logos, aggregate ratings or anonymous success claims as proof.
Price expectations are addressed
Test: The page should give a price, range, starting point or the factors that determine a quote.
If it fails: Explain scope and exclusions so serious prospects can self-qualify before enquiring.
Forms ask only useful questions
Test: Complete the form on mobile. Identify fields that do not affect qualification, routing or the first reply.
If it fails: Remove unnecessary friction while keeping the details needed for a useful response.
Every enquiry route is tested end to end
Test: Submit the form, click phone and email links, confirm the thank-you state and verify the owner actually receives the notification.
If it fails: Add monitoring and a fallback route; page appearance alone does not prove delivery.
The response promise matches operations
Test: Check stated turnaround, opening hours and follow-up wording against what can consistently be delivered.
If it fails: Use a realistic promise and send an immediate acknowledgement that sets the next step.
Section 5
Measurement and prioritisation
Finish the audit with evidence, ownership and a small sequence of changes that can be verified.
Search Console and analytics are connected
Test: Confirm ownership, data collection, timezone and access. Keep branded and non-branded discovery separate.
If it fails: Fix verification and tracking before making traffic conclusions from guesswork.
Meaningful conversions are recorded
Test: Check form success, qualified calls, booked consultations and purchases rather than counting button clicks as leads.
If it fails: Use thank-you or server-confirmed events and test them in the analytics debugger.
Lead source survives the hand-off
Test: Follow a test enquiry into email, CRM or the lead store. Confirm landing page and campaign data remain attached.
If it fails: Store source data with the lead so marketing can be tied to accepted work.
A baseline is dated
Test: Record indexable pages, impressions, clicks, qualified enquiries, conversion rate and key performance evidence before changes.
If it fails: Annotate deployments and compare like-for-like periods rather than relying on memory.
Actions are ranked by dependency
Test: Access and indexing come before snippets; correct intent comes before more copy; working forms come before buying traffic.
If it fails: Select the three highest-confidence actions and state expected evidence, owner and effort.
Every fix has a validation method
Test: Define what proves implementation and what later outcome would support the hypothesis.
If it fails: Re-crawl technical fixes immediately, then wait for enough search or conversion data before judging impact.
Primary references
The checklist is grounded in observable website behaviour and current search documentation. Start with Google’s guidance on robots.txt, canonical URLs, Core Web Vitals and people-first content.
Checklist questions
How do I use this website audit checklist?
Start with the pages responsible for enquiries, not every URL. Work through the sections in order, record evidence beside each failed check and finish with three actions ranked by dependency, confidence, likely impact and effort.
Can this checklist replace a professional SEO audit?
It can identify many common problems and improve a brief for a developer or auditor. It cannot replace Search Console interpretation, a complete crawl, template sampling, analytics investigation or specialist review on a large or technically complex website.
What should I fix first after a website audit?
Fix anything preventing access, indexing or enquiries first. Then correct page purpose and misleading claims before polishing snippets or publishing more content. A small verified sequence is safer than changing everything simultaneously.
How often should I repeat the checklist?
Repeat the relevant sections after a launch, migration or material template change. For a stable small website, monitor forms, indexing and performance continuously and run a broader review when evidence changes rather than simply on a calendar.
Does passing every check guarantee traffic?
No. The checklist improves eligibility, clarity, usability and measurement. Search visibility still depends on demand, competition, relevance, external authority and search-system decisions outside the website owner’s control.
Want the three checks that matter on your site?
Send the URL and the result you need. Keiron will review it by hand and return a six-page audit with the highest-priority actions first.
Request the free audit