If you have dozens of posts and no clear next step, start with one commercially relevant topic group. You can make a useful decision from a carefully chosen set of 10–20 URLs before committing to a site-wide rewrite.
Choose the first batch around a business question
Start with a service or product you actually want to sell. Gather its main landing page, supporting articles and any older pages answering similar questions. This shows whether the content forms a useful route to the offer.
Include a mix: a page already attracting relevant searches, a promising page with limited clicks, a page that has declined and one that seems to overlap with another. A list made only from the worst traffic numbers can miss your best improvement opportunity.
Keep a whole-site URL inventory in the background. The first batch is a working scope, not permission to ignore important pages elsewhere. If you are deciding what to publish next rather than evaluating existing URLs, use a research-led content brief instead.
Put the question, evidence and business role in one row
Record the canonical URL, page type, intended reader, question it answers, target market and relevant offer. Add the available search and business evidence with its reporting period.
Useful inputs include Search Console queries and landing pages, analytics, internal links, backlinks and actual inquiry or sales records. Label missing data as unknown. A new page without enough history is not the same as a mature page losing demand.
Compare like with like: the same country, device group and appropriate dates. A seasonal buying guide may need a year-over-year view. Check recent URL or tracking changes before blaming its copy. Search Console documents that query rows are incomplete and much performance data is assigned to canonical URLs; an export is not a complete record of every search.
AI-assisted drafts need the same useful-page standard plus a close check of their claims and sources. The AI content audit shows an annotated revision decision.
Read the page and the results your customer sees
A keyword tool can identify a topic worth investigating. It cannot establish whether your answer is useful. Open the page on a phone and follow the customer's task from the headline to the next step.
Then review relevant competing results for the intended market. Note what they do well: a clear explanation, a helpful comparison, original photos or an example that resolves a buying objection. Look for what your page can explain better, not just headings to copy.
- Does the search call for advice, a product collection or a service provider?
- What important decision is still unanswered?
- Which claims need a source, an expert check or better evidence?
- Can the reader find the relevant service without being pushed toward an unrelated sale?
Google's people-first content questions are useful prompts for evaluating substance. There is no required word count to reach.
Three pages can need three different decisions
The following is an illustrative review of a service-business content group. It shows the reasoning a useful audit should expose; these are not findings from a client site.
Choose a small number of changes the team can finish
Put an important, fixable obstacle ahead of a cosmetic rewrite. A useful article with a missing buying answer may deserve attention before an unrelated high-volume topic. A blocked page needs a technical investigation before an editor spends hours rewriting it.
For each recommendation, specify the URL, missing information, proposed change, supporting evidence, owner and completion check. Separate a confirmed problem from a question that still needs investigation.
Some work is a focused edit; some needs fresh research or a substantial rewrite. If two pages appear to compete, use the keyword cannibalization diagnosis before merging them. For a single page's refresh-versus-new decision, see when to update existing content.
Ask for an improvement plan, not just an inventory
An inventory tells you what exists. A useful content audit also explains what to do with it. Agree the URL count, topic depth, data access and whether the work includes briefs, rewritten copy or publication.
| Deliverable | What makes it useful |
|---|---|
| URL decision list | A reason to keep, improve, separate or combine each reviewed page |
| Priority notes | The reader or business problem being addressed and the dependency on other work |
| Editing brief | Specific information, examples, evidence and links needed on the selected pages |
| Implementation record | What was actually changed, by whom and when |
| Follow-up plan | What to check after publication and which results need more time |
A full technical audit, expert legal or medical review, and content implementation are different scopes. Confirm the parts you need rather than assuming one file includes everything.
Check the work before judging the result
First check the published page: does it answer the intended question, retain useful material, load correctly and point to the right next step? Verify links, canonicals and redirects when URLs changed.
Then monitor the affected page group under comparable conditions. Track relevant search queries and qualified inquiries where you can. Keep a dated change log so a seasonal shift, promotion or site release is not automatically credited to the edit.
I can help research and improve a defined set of pages through content and on-page SEO support. Send the site and the topic group you want to improve; we can decide whether the first job is a review, a rewrite or a wider plan.
Before you change existing content
- Match each content URL to its search intent, qualified leads, internal links, and current Search Console performance.
- Separate content problems from indexing problems by comparing indexability, the user-declared canonical, and Google’s selected canonical.
- Assign each keep, update, merge, or remove decision to an editor or developer, with affected URLs and a validation step.
Questions before a content audit
Should we audit existing content before paying for new articles?
Start with an audit when you already have a content library but cannot explain which pages generate qualified inquiries or support sales. It can reveal outdated pages to improve, overlapping topics to consolidate, and genuine gaps that need new content. If the site has little relevant content, search-intent research and a focused publishing plan may be the better starting point.
Does low traffic mean a page should be deleted?
No. A niche service page may attract few visits but bring valuable estimate requests or support sales conversations. Review conversions, relevant backlinks, seasonal demand, and indexing before removing it. Use a 301 redirect when a genuinely equivalent replacement exists; a permanently removed page without a replacement can return a 404 or 410.
Can our writer implement the audit, or will we need a developer?
An editor can handle search intent, outdated facts, headings, and contextual internal links. A developer may be needed for redirects, canonical rules, templates, or rendering issues. Ask for URL-level recommendations, acceptance checks, and a clear split between editorial and technical work so your team knows exactly what to implement.