An SEO Audit Template You Can Reuse: Structure, Scoring and How to Present It

An SEO audit template should contain five sections, a finding format and a priority scheme. The sections are crawl and index, content, links, local presence and AI visibility. The finding format has fields for the issue, the evidence, the affected pages, the likely impact, the effort, the owner and the status. The priority scheme turns those fields into an order of work.

The template is a structure to fill in from your own data, not a list of verdicts. Every row needs evidence you can reproduce, such as a Search Console report, a crawl export or a screenshot of a results page. An audit with no evidence is an opinion, and it cannot be reviewed or repeated next quarter.

Decide before you start who the audit is for. An audit for a developer needs URLs, status codes and reproduction steps. An audit for an owner needs consequences in plain language, such as which services cannot be found by customers searching for them. The same findings can serve both if the template has a plain-language summary column next to the technical detail.

The crawl and index section checks whether search engines can reach, render and store your important pages. Use the Search Console page indexing report to see which URLs are indexed and why others are not, and URL Inspection to see how Google fetched and rendered a specific page. Then run a crawl of your own to find broken links, redirect chains, missing canonicals and pages with no internal links.

Compare three lists: pages you want indexed, pages that are indexed, and pages in your sitemap. The gaps tell you where to look. Pages you want that are not indexed are the top priority. Pages that are indexed but should not be, such as internal search results or duplicates, usually need a canonical or noindex decision. Record robots.txt rules too, because a single disallow line can hide a section that took months to write.

Also record how the site behaves on a phone, since Google primarily uses the mobile version of a page for indexing. Compare content, headings, links and structured data between the mobile and desktop versions, and note anything present on one but missing from the other. Where you rely on JavaScript to show content, confirm that the rendered HTML in URL Inspection contains it.

Audit content by matching each page to a reader question and checking whether the page answers it better than what ranks. Export queries and pages from the Search Console performance report, find pages with impressions but few clicks, and read the results page for those queries yourself. Mark each page keep, improve, merge or remove. Google's guidance on helpful content asks whether a page gives original, complete information for the people it serves, which is a workable test for the keep or improve decision.

Audit links in two directions. Internally, check that important pages are reachable from navigation and related pages with descriptive anchor text. Externally, list the sites that link to you and note whether any look purchased or manipulative, since Google's spam policies cover link schemes. For local, compare your business name, address, phone number, categories and hours across your Google Business Profile, your website and major directories, and note review volume and response habits. Skip the local section if you have no physical service area.

For content, add a column for the primary reader question and one for whether the page actually answers it in its first paragraph. This catches the common pattern of a long page whose useful answer sits at the bottom. For links, a short list of the five or six best pages that earn mentions is usually more useful than a full backlink export no one will read. Local and AI sections should name an owner as well. Local findings often sit with whoever manages the Business Profile and reviews, while AI visibility findings usually need an editor, because the fix is clearer and better-sourced pages rather than a technical change.

AI visibility belongs in the audit as a short check, not a separate programme. Google states that the same fundamentals apply to its AI features in Search as to ordinary results: pages must be indexed and eligible to appear with a snippet, and there are no extra technical requirements. So the section asks whether your key pages meet those conditions, then whether they answer questions directly and carry evidence worth citing.

Run a manual sample. Take ten questions a customer would ask, put them to the main assistants and to Google, and record which sources are cited, whether your pages appear, and whether what is said about your business is accurate. Treat the result as a snapshot, not a ranking, because answers change between runs. Note factual errors about your business; those are fixed by clarifying your own pages, not by tricks. Do not add structured data types your pages cannot honestly support just to look more machine-readable.

Score each finding on impact and effort, then sort into four groups. High impact and low effort goes first. High impact and high effort becomes a planned project. Low impact and low effort is a filler task for a quiet week. Low impact and high effort is recorded and declined. Impact should be judged by whether the affected pages earn enquiries and whether the problem blocks visibility, not by how many pages the crawler flagged.

Avoid a single numeric site score. A composite number invites argument about the formula and tells nobody what to do on Monday. If your stakeholders want a summary measure, report counts in each priority group and a status for each section: blocking, needs work, or sound.

Present the audit as a one-page ordered action list followed by evidence. Lead with the three or four changes that matter most, say why each matters in business terms, name an owner, and give a date for review. Put the full findings table in an appendix for the people who will implement it. A reader who only sees the first page should still know what to do.

Keep the language plain and the claims modest. Say what you observed, what you think it means, and how you would confirm the effect, usually by comparing Search Console data before and after the change. Re-run the audit on a schedule, and keep the same template so you can see which findings were closed and which keep returning.

A closing section titled "What we did not check" is worth including. If you could not access server logs, or skipped a competitor comparison, say so. It tells the reader where the audit's limits are and prevents a clean report from being read as a clean bill of health.