Website audit questions, answered clearly

A website audit is a structured review of how well a site can be found, understood and used. A useful audit checks technical crawl signals, search intent, page quality, mobile experience, trust, conversion paths and measurement. It should finish with a prioritised action list, not a dump of warnings.

Different audits answer different questions. A technical SEO audit asks whether search engines can crawl and interpret the site. A conversion audit asks why visitors do not call or enquire. A growth audit joins those views so the recommended fix matches the actual business problem.

At minimum: indexability, canonical URLs, redirects, sitemaps, robots directives, page titles, internal links, mobile usability, Core Web Vitals, contact paths and analytics. For a local service business it should also examine service-page coverage, Google Business Profile alignment, reviews and local proof.

The report should identify evidence, likely impact, effort and ownership for each recommendation. Our audit contents page shows the six-page format used here.

An SEO audit concentrates on organic visibility: crawling, indexing, content relevance, internal links, structured data and authority signals. A website audit can be broader, adding user experience, accessibility, conversion, tracking and commercial messaging.

If the complaint is “we are not showing on Google”, begin with SEO and indexing evidence. If the site gets visits but no enquiries, an SEO-only report may miss the real leak. The audit brief should state the business outcome before anyone opens a crawler.

Automated tools are excellent at repeatable checks: missing tags, response codes, page weight, broken links and measurable performance. They cannot reliably decide whether a page answers the searcher, whether the offer is believable, or which issue deserves the next hour of work.

Use automation to collect evidence and a human review to interpret it. A score of 82 is not a strategy. The important question is which failed check is suppressing qualified traffic or enquiries.

A focused review of a small service-business site can often be completed in one or two working days. Large ecommerce, multilingual or JavaScript-heavy sites need longer because templates, filters, product feeds and rendering states must be sampled separately.

Speed should not remove verification. A credible auditor checks important findings in the browser and source, distinguishes site-wide issues from one-page issues and records uncertainty where analytics or Search Console access is missing.

The usual causes are lack of indexation, weak demand targeting, insufficiently useful pages, poor internal discovery or too little external authority. A technically perfect site can still receive no traffic if its pages do not answer searches people make or stronger sources deserve the result.

Check Search Console impressions before guessing. Zero impressions suggests discovery, indexing or relevance. Impressions without clicks point toward ranking position, title and snippet fit. Visits without enquiries point toward intent, trust or conversion.

First separate branded and non-branded searches. A new site may appear for its exact name but not competitive service queries. Confirm the page returns 200, has no noindex, uses a self-consistent canonical, appears in the XML sitemap and is linked from another crawlable page.

Then use Search Console URL Inspection. Google explains that sitemaps and internal links support discovery, but neither guarantees indexing or ranking. The page still needs distinct, useful content and a reason to be selected over alternatives.

Traffic can be informational, poorly targeted or unable to convert. Compare landing page, query intent and action. A visitor reading a glossary definition is not equivalent to someone searching for a supplier. On commercial pages, check whether price expectations, proof, location, response time and the next step are clear on mobile.

Also test the plumbing: submit every form, click every phone link and confirm delivery. More traffic amplifies both a working funnel and a broken one.

Search engines normally choose a representative canonical when pages are very similar; it is not automatically a penalty. The practical problem is dilution. Near-identical pages can compete for the same intent, split internal signals and give Google little reason to index every version.

Consolidate pages that answer the same question. Keep separate URLs only where the audience, evidence and answer are materially different. Canonicals are not a substitute for deciding which page should exist.

Use a dependency order. Fix access and indexability before snippets; fix the correct page and intent before adding more copy; fix measurement before judging conversion. Then prioritise by expected business impact, confidence and effort.

A missing meta description is rarely more urgent than a noindexed service page, a redirecting enquiry page or a form that does not deliver. A good report explains this order rather than giving every warning the same colour.

Google includes Core Web Vitals within its broader page-experience considerations, but relevance and helpfulness remain fundamental. The current metrics are LCP for loading, INP for responsiveness and CLS for visual stability. Google recommends aiming for LCP within 2.5 seconds, INP below 200 ms and CLS below 0.1 at the 75th percentile.

Improve poor real-user performance, but do not trade away useful content or conversion clarity to chase a perfect lab score.

Pricing depends on site size, access and depth. A free review is usually a focused triage; a paid technical or commercial audit may range from several hundred pounds for a small site to several thousand for ecommerce or complex platforms.

Compare scope rather than page count. Ask whether crawling, Search Console evidence, analytics, competitors, templates, conversion testing and an implementation call are included. A long generic PDF can be less valuable than a short report that identifies the right three actions.

A public first pass needs only the URL and the outcome you want. Deeper diagnosis benefits from read-only Search Console and analytics access, because impressions, clicks, landing pages and conversions reveal whether the issue is discovery, demand or on-site behaviour.

Do not share passwords. Add users through the platform, grant the minimum read permission required and remove access after the work if it is no longer needed.

Yes. Record current organic landing pages, queries, backlinks, conversions, titles, canonicals and redirects before changing URLs or content. Otherwise a redesign can remove the pages and signals that were already working.

The website redesign SEO checklist covers the migration controls. Design decisions should preserve proven intent and improve weak journeys, not reset the site blindly.

Run targeted checks after deployments, migrations or sharp traffic changes. A broader review every six to twelve months is reasonable for a stable small site; frequently changing ecommerce or publishing sites need automated monitoring and more regular sampling.

Do not rerun a giant checklist on a calendar without comparing outcomes. Monitor index coverage, conversions, performance and broken journeys continuously, then investigate when the evidence changes.

Confirm the baseline, assign each action, implement the smallest high-confidence fixes first and annotate the change date. Re-crawl affected URLs, test forms and redirects, then watch Search Console and analytics over an appropriate period.

Separate validation from outcome. A correct canonical can be verified immediately; ranking and conversion impact needs enough impressions and visits to judge. Keep a decision log so improvements are not reversed later.