Key takeaways
- An enterprise SEO audit works by template and page type: find the few templates that produce most URLs, sample each one, and fix the template.
- Start with an inventory of URLs per page type from sitemaps, a crawl, Search Console and server logs, before you look at a single issue.
- Search Console’s Page indexing report lists at most 1,000 example URLs, so split your sitemaps by page type to see each section clearly.
- Group findings by root cause and owner. Ten findings with clear owners beat four hundred warnings nobody reads.
- The deliverable is a ticket list engineers accept, with pages affected, the fix, acceptance criteria and a priority.
An enterprise SEO audit fails in a predictable way: someone crawls a million URLs, exports every warning, and hands over a spreadsheet so large that nobody acts on it. Six months later the same problems are still there. On a very large site the job isn’t to find every issue. It is to find the few causes behind most of them, and get them fixed by the teams who own them.
I have audited sites of every size in more than 16 years of SEO. This is the method I use on large ones: audit by template and page type, sample instead of trying to review everything, use Google’s own data to confirm what matters, and turn findings into tickets.
Why page-by-page audits fail on large sites
On a 50-page site you can look at every page. On a site with hundreds of thousands of URLs, the same approach produces three problems:
- Volume hides causes. Ten thousand “missing meta description” warnings are usually one template bug, reported ten thousand times.
- Crawls never finish cleanly. Faceted navigation, parameters and calendars can generate URLs faster than any crawler can fetch them.
- Nobody owns a spreadsheet. Findings that aren’t tied to a template and a team stay in the report.
Google’s own bar for an audit is short: “An SEO audit should be about giving you realistic estimates of improvement, and an estimate of the work involved.” A spreadsheet of warnings meets neither half.
On a large site, a finding without a template and an owner is just a number.
Step 1: Build an inventory of URLs by page type
Before looking for issues, find out what the site actually is. I build one table: every page type, how many URLs it has, how many are in the sitemap, how many Google has indexed, and who owns the template.
| Source | What it tells you |
|---|---|
| XML sitemaps (split by page type) | What the site says should be indexed |
| A crawl (limited depth, parameters controlled) | What internal links expose |
| Search Console Page indexing report | What Google has indexed, and why not |
| Search Console Crawl Stats | What Googlebot actually requests, and how the server responds |
| Server logs (if available) | Every request, including URLs nobody knew existed |
| Analytics landing pages | Which page types bring organic visits and revenue |
The mismatches are where audits find value: page types with many URLs but few indexed, sections Googlebot crawls heavily that bring no traffic, or valuable pages missing from sitemaps.
Use Search Console the way large sites need to
The Page indexing report “shows how many URLs on your site have been crawled and indexed by Google”, but “the list of example URLs in the report is limited to 1,000 items”. On a large site, that limit makes site-wide views almost useless for diagnosis. The fix is simple: submit separate sitemaps per page type and use the “All submitted pages” filter, so each section has its own view.
Read the report with Google’s own expectation in mind: “Don’t expect every URL on your site to be indexed.” The goal it gives is “to get the canonical version of every important page indexed.” Reasons like “Crawled – currently not indexed”, “Discovered – currently not indexed” and “Duplicate without user-selected canonical” are only problems when they hit the page types you care about.
The Crawl Stats report is “aimed at advanced users”, and Google says “If you have a site with fewer than a thousand pages, you should not need to use this report.” Large sites do need it. Its breakdowns by response type, file type, crawl purpose (discovery or refresh) and Googlebot type show where crawling goes, and host status shows whether Google “encountered availability issues when trying to crawl your site.”
Step 2: Sample each template instead of reviewing every page
Once you know the page types, sample them. For each important template I review a few dozen URLs chosen on purpose: the best performers, the worst, the newest, the oldest, and some at random. Then I check the things that templates control:
- Status code, canonical tag and meta robots;
- Title, main heading and indexable body content in the rendered HTML;
- Internal links in and out (navigation, breadcrumbs, related items, pagination);
- Structured data, if the template uses it;
- Page speed on a representative page.
If the sample shows the same problem on most URLs of a template, it is a template problem, and one fix covers every page. If it only shows on a few, it is data or content, and the fix sits with a different team.
Step 3: Follow the crawl, not just the pages
On large sites the biggest wins are often in URLs that shouldn’t exist. Google’s crawl budget guide recommends managing your URL inventory, consolidating duplicate content, blocking crawling of unimportant URLs with robots.txt, returning 404 or 410 for removed pages, eliminating soft 404s, keeping sitemaps up to date and avoiding long redirect chains. In an audit I look for:
- Parameter and filter combinations that create near-infinite URLs;
- Duplicates across sorting orders, tracking parameters, upper and lower case, trailing slashes;
- Soft 404s, such as empty category or search pages that return 200;
- Redirect chains left behind by earlier migrations;
- Noindexed sections that are still crawled. Google says it “will still request, but then drop the page”, so noindex doesn’t save crawling.
Step 4: Group findings by root cause and owner
This is the step that decides whether the audit gets used. I group every finding by the cause and the team that can fix it, not by the crawler’s category. Many tools report “duplicate titles”, “thin content” and “canonical issues” separately when all three come from the same faceted category template.
Step 5: Turn findings into tickets teams will ship
The deliverable of an enterprise SEO audit isn’t a report. It is a list of tickets. Each one needs:
| Field | Example |
|---|---|
| Finding | Category pages with filters applied are indexable and duplicate the unfiltered page |
| Pages affected | Page type and URL count, with sample URLs |
| Evidence | Page indexing reasons, crawl data, rendered HTML |
| Fix | What changes in the template or rule |
| Acceptance criteria | What QA checks on staging before release |
| Owner | The team that owns the template |
| Priority | Based on traffic and value of the affected page type, and effort |
Enterprise SEO audit deliverables
- URL inventory by page type, with indexed counts and owners.
- Sample review per important template.
- Crawl findings: parameters, duplicates, soft 404s, redirect chains.
- Findings grouped by root cause and owner.
- Ticket list with evidence, fix, acceptance criteria and priority.
- A short summary for leadership: the five changes that matter most and the effort each needs.
If you want to see what a smaller-site audit covers, our article on what a real SEO audit includes walks through it. For migrations, the audit should also feed rank tracking after the move.
How to do an enterprise SEO audit: access and timing
Ask for read access first. Google advises that at the audit stage you “only grant read access to Search Console (at this stage, don’t grant them write access).” For a large site I also ask for sitemaps, a sample of server logs if they exist, analytics read access and a list of templates and their owners.
Timing depends on the number of page types far more than the number of URLs. A site with a million URLs and eight templates can be quicker to audit than one with fifty thousand URLs and forty templates. I scope audits by the hour, so the scope follows the site’s actual complexity. If traffic is already falling and you need a cause fast, see the checks in why organic traffic is decreasing.
FAQ
How is an enterprise SEO audit different from a normal SEO audit?
It works by template and page type instead of page by page, relies on sampling and on Google’s own data (Page indexing and Crawl Stats) more than on one full crawl, and delivers tickets for named teams rather than a single report.
How do you do an enterprise SEO audit without crawling the whole site?
Build an inventory of page types first, crawl with limits on depth and parameters, sample each template, and use sitemaps split by page type so Search Console shows each section separately.
How long does an enterprise SEO audit take?
It depends on the number of templates, markets and data sources, not just the number of URLs. A site with a few templates can be audited much faster than a smaller site with dozens of different page types.
What should an enterprise SEO audit deliver?
An inventory of page types, findings grouped by root cause and owner, and a prioritised ticket list with evidence, fixes and acceptance criteria, plus a short summary of the most important changes for leadership.
If your last audit produced a spreadsheet nobody acted on, send me a list of your page types and I'll tell you how I would audit the site instead.
Eugen Platon · SEO consultant with 16+ years in SEO and digital marketing · free 10-minute first call