The complete website audit checklist for agencies (2026)

Every agency that audits more than five client sites a month eventually builds a homegrown dashboard. Most of them regret it. The spreadsheet-plus-Screaming-Frog-plus-Looker-Studio stack works until it doesn't — which is usually around client number eight, when the formulas break, the exports stop matching, and someone overwrites the template tab.

This checklist is built for agencies specifically. Not in-house teams with one site, not solo consultants with two. Agencies — the people managing a portfolio of client websites, each with different stacks, different problems, and different stakeholders who need reports they'll actually read.

If you've done this before, skip straight to the checklist. If you want the workflow context first, start here.

Why agency audits are different

An in-house team audits one site deeply. An agency audits many sites at consistent depth. The difference is operational, not technical. You need:

• Repeatable process — the same methodology applied to every client, regardless of stack
• Prioritized output — clients don't want 300 findings; they want the 10 that matter
• Shareable deliverables — reports that make sense to someone who doesn't know what a canonical tag is
• Ongoing monitoring — not a one-time audit, but continuous visibility into site health

The checklist below addresses all four. It's ordered by audit phase, not by importance — because the order you check things matters when you're doing this across multiple sites in a week.

Phase 1: Crawl and inventory (items 1–5)

Before you evaluate anything, you need a complete picture. A crawl gives you every URL, its status, and its metadata in one pass.

1. Run a full site crawl

Crawl the entire site from the root URL. Capture every page the crawler can reach, including pages not in the navigation. Note the total page count — this is your baseline.

2. Check crawl coverage against the XML sitemap

Compare crawled URLs against the XML sitemap. Pages in the sitemap but not discovered by the crawler are likely orphaned or blocked. Pages discovered by the crawler but missing from the sitemap may need to be added.

3. Record HTTP status codes

Flag all non-200 responses. Separate into categories:

• 301/302 redirects — are they intentional? Do redirect chains exist?
• 404 errors — are these pages still linked to internally?
• 5xx server errors — are these intermittent or persistent?

4. Capture page-level metadata

For every URL, record: title tag, meta description, H1, canonical URL, robots directives, word count, and internal link count. This becomes the foundation for items 6–20.

5. Document the tech stack

Note the CMS, framework, hosting provider, and CDN. This context matters for interpreting audit results — a missing meta description on a Payload CMS site has a different root cause than on a WordPress site.

Phase 2: Indexability and crawlability (items 6–10)

These are the foundational issues. A page that can't be crawled or indexed can't rank, regardless of its content quality.

6. Verify robots.txt is not blocking important content

Check that robots.txt exists, is valid, and isn't accidentally disallowing critical sections. Pay special attention to staging subdomain rules that may have been copied to production.

7. Identify noindex pages that should be indexed

Filter your audit data to show all pages with noindex directives — whether set via meta tag, HTTP header (X-Robots-Tag), or robots.txt. Cross-reference with Google Search Console impressions. Any noindex page receiving search impressions is a red flag: Google knows about it, users are looking for it, but you're telling Google not to show it.

8. Audit canonical tags

Every indexable page should have a self-referencing canonical tag. Check for:

• Missing canonicals
• Canonicals pointing to non-200 URLs
• Canonicals pointing to noindexed pages
• Conflicting canonicals across HTTP/HTTPS or www/non-www variants

9. Check for redirect chains and loops

Any redirect path longer than one hop wastes crawl budget and dilutes link equity. Map all redirect chains and flatten them to single-hop 301s.

10. Validate the XML sitemap

The sitemap should contain only indexable, canonical, 200-status URLs. Remove anything that redirects, returns a 404, is noindexed, or has a canonical pointing elsewhere.

Phase 3: On-page SEO (items 11–16)

With indexability confirmed, evaluate the on-page signals that influence rankings.

11. Audit title tags

Check for:

• Missing titles — pages without a <title> element
• Duplicate titles — multiple pages sharing the same title
• Truncated titles — over 60 characters, cut off in SERPs
• Keyword-stuffed titles — readable by humans, not just crawlers

12. Audit meta descriptions

Check for missing, duplicate, and over-length descriptions (aim for 120–155 characters). While Google often generates its own snippets, well-written descriptions still improve click-through rates when they're used.

13. Validate heading hierarchy

Every page should have exactly one H1 that describes the page content. The heading hierarchy should descend logically — no H3 under an H1 without an H2 between them. Check for missing H1s, duplicate H1s, and H1s that duplicate the title tag verbatim (wasted real estate).

14. Check for thin content pages

Pages with fewer than 300 words of substantive content rarely rank and can drag down overall site quality. Flag pages under that threshold, but evaluate each in context — a contact page doesn't need 1,000 words.

15. Identify duplicate and near-duplicate content

Find pages with identical or near-identical body content. These compete with each other in search results and should be consolidated (canonical to one, redirect the others) or differentiated.

16. Audit internal linking

Check internal link distribution. Every important page should be reachable within three clicks from the homepage. Look for:

• Orphan pages — pages with zero internal links pointing to them
• Dead-end pages — pages that link out to nothing
• Excessive link concentration — the homepage linking to everything while deeper pages link to nothing

Phase 4: Performance and technical health (items 17–21)

Performance issues affect both rankings and user experience. Clients notice these immediately because they experience them.

17. Run site-wide Lighthouse audits

Don't test just the homepage. Run Lighthouse across the entire site — or at least across every template type. The homepage might score 95 while product pages score 40. Average scores mislead; distribution matters.

18. Evaluate Core Web Vitals

Check the three CWV metrics across the site:

• LCP (Largest Contentful Paint): target under 2.5 seconds
• INP (Interaction to Next Paint): target under 200 milliseconds
• CLS (Cumulative Layout Shift): target under 0.1

Use both lab data (Lighthouse) and field data (CrUX via Search Console) when available. Lab data tells you what could happen; field data tells you what does happen.

19. Test mobile rendering

Verify the site renders correctly on mobile devices. Google uses mobile-first indexing, so mobile rendering issues directly affect rankings. Check for content parity between mobile and desktop — anything hidden on mobile may be deprioritized.

20. Check for broken resources

Broken images, missing CSS files, and failed JavaScript loads degrade both performance and user experience. A crawl surfaces these, but manual spot-checking on key pages catches rendering issues that crawlers miss.

21. Verify HTTPS and mixed content

Every page should load over HTTPS. Check for mixed content warnings — HTTP resources loaded on HTTPS pages. These trigger browser warnings and erode trust.

Phase 5: Reporting and monitoring (items 22–25)

The audit is only useful if it translates into action. For agencies, that means deliverables clients understand and systems that catch regressions.

22. Prioritize findings by impact

Not every issue deserves immediate attention. Categorize findings into three tiers:

• Critical — directly blocking rankings or causing user-facing errors (broken pages, noindex on key content, server errors)
• Important — degrading performance or missing optimization (thin content, missing metadata, slow pages)
• Advisory — best practice improvements (heading hierarchy, internal link optimization, image alt text)

Present the critical tier first. Always.

23. Create a shareable report

Static PDF reports get downloaded and forgotten. What works better is a report URL the client can bookmark — something that shows the current state, not the state at the time of the last export. Include only the findings relevant to the client's goals. A 200-page technical report impresses nobody.

In Evergreen, you can generate a shareable report URL that shows the audit results in a client-friendly view. The client sees the data without needing an account, and the report updates as you re-crawl the site.

24. Set up continuous monitoring

A one-time audit captures a snapshot. Continuous monitoring catches regressions — the new deployment that accidentally noindexed the blog, the CDN change that doubled LCP, the intern who deleted the sitemap. Set up automated crawls on a daily or weekly cadence for every active client.

25. Document the audit methodology

For every client engagement, note the tools used, the crawl date, the scope (full site vs specific subdirectory), and any known limitations. This protects the agency when clients ask "why didn't you catch this?" six months later. It also makes the next audit faster.

The agency workflow: from prospect to monitoring

Here's how this checklist fits into a typical agency engagement:

New client onboarding. Crawl the site within the first 48 hours. Run items 1–21 to establish a baseline. Deliver the initial report (items 22–23) within the first week.

Monthly maintenance. Re-crawl on a set cadence. Review the delta — what changed since last month? Focus client communication on changes, not the full audit.

Quarterly deep reviews. Run the full 25-point checklist end to end. Compare against the baseline from onboarding. Highlight trends: is the site getting healthier or degrading?

Before major changes. Any migration, redesign, or replatform gets a pre-change audit (the baseline) and a post-change validation audit within 48 hours of launch.

How Evergreen handles this

Evergreen is built for the multi-site workflow agencies need. Add up to 10 client projects on the Pro plan ($49/month), each with its own crawl, audit table, visual sitemap, and shareable report URL.

The audit table surfaces every item on this checklist — missing metadata, indexability issues, broken links, thin content — in a sortable, filterable view. Filter to critical issues, share the filtered view with the client, and move on to the next site.

The visual sitemap gives clients something they can actually understand. Instead of explaining internal link depth with a spreadsheet, you show them their site structure in a diagram where problems are visible.

Shareable reports generate a public URL that clients can bookmark. No login required, no PDF to lose. The report shows the current state of the site, updated with each new crawl.

Daily syncs (Pro tier) catch regressions automatically. When the client's team deploys a change that breaks something, you see it the next morning — not the next quarter.

The entire flow — crawl, audit, report, monitor — happens in one tool, across all your client sites, without switching between Screaming Frog exports, Google Sheets, and Looker Studio.

Manage 10 client sites with daily syncs for $49/mo → Start free

Frequently asked questions

How long should a full website audit take?

For a site under 500 pages, a thorough first-time audit takes 2–4 hours including the report. Subsequent audits take less because you're reviewing changes, not starting from scratch. For larger sites (5,000+ pages), budget a full day for the initial audit and automate the recurring checks.

What's the minimum viable audit for a new client?

Items 1–10 (crawl and indexability) plus items 17–18 (performance). These catch the issues that directly block rankings. You can run the full 25-point checklist in the second month once you understand the client's priorities.

How do you explain audit findings to non-technical clients?

Focus on outcomes, not terminology. Instead of "17 pages have conflicting canonical tags," say "17 pages are sending confusing signals to Google about which version to show in search results — this means none of them rank as well as they should." Pair every finding with a recommended action and its expected impact.

Should agencies use free tools or paid audit platforms?

Free tools work for one-off projects. Paid platforms pay for themselves at three or more active clients — the time savings on reporting, monitoring, and multi-site management cover the cost. The real cost isn't the subscription; it's the hours spent stitching exports together.

Your next step: crawl your first client site in 60 seconds → Create free account