Most content audits produce a spreadsheet of every URL on the site, a color for each one, and a recommendation to "refresh" half of them. A month later the spreadsheet is still open and nothing has changed, because nobody could tell which rows mattered.
I run audits with a narrower question: where is this site already close to earning more, and what is the smallest change that gets it there? The answer almost never involves writing new pages. It usually sits in pages that already rank, already get seen and already have a reason to exist, but are leaving something on the table.
This is the method. It needs Search Console, your analytics, and a spreadsheet. It does not need a paid tool.
What this audit is, and what it is not
This is not a pruning exercise. Deciding which pages to merge, redirect or delete is a different job with different risks, and I cover it in content decay: why deleting posts made my sites stronger. Run that first if your archive is large and old.
This audit looks only for upside. Every row it flags should come with a specific change and a reason to believe the change will be worth more than it costs.
Step one: build one sheet from two sources
From Search Console, open the Performance report, switch the table to Pages, set the date range to the last twelve months and export it. You get four numbers per page. Google defines them precisely, and the definitions matter for what follows: clicks are clicks from Search results to your site, impressions are how many times your site appeared in results, CTR is clicks divided by impressions, and position is the average position of the topmost result from your site (Search Console Help).
Two limits to know before you trust the export. The interface exports at most 1,000 rows, and some queries are left out entirely to protect user privacy (Google Search Central). For a site with more than a thousand pages that get impressions, pull the data through the Search Analytics API or the Looker Studio connector instead, which the same post says can return up to 50,000 rows per day per site per search type.
From your analytics, export conversions by landing page for the same period. A conversion is whatever earns money for this site: a sale, a lead form, a booking, a signup that turns into revenue.
Join the two on the page URL. You now have one row per page with clicks, impressions, CTR, position and conversions. That sheet is the audit. Everything else is filtering it.
Step two: filter for the six patterns that hide money
1. Pages on the edge of page one
Filter for pages with an average position roughly between 8 and 20 and meaningful impressions. These pages already have Google's attention. They are being shown, just not high enough to be clicked.
Do not act on the page average alone. Average position blends every query a page appears for, so a page can sit at 12 on average while ranking near the top for one query and far down for ten others. Open the page in the Performance report, switch to Queries and find the specific searches where it is close. The fix is usually focused: answer the query more directly in the opening section, add the subtopic the top results cover and yours does not, and point two or three relevant internal links at it from stronger pages. Internal linking architecture covers how to choose them.
2. Pages that are seen but not chosen
Sort the pages already in the top few positions by CTR, lowest first. A page that ranks well and still gets few clicks for its impressions has a presentation problem, not a ranking problem.
Remember what you actually control. Google uses a number of different sources to determine a result's title link (Google Search Central), and it primarily uses the content on the page to build the snippet, using the meta description when that describes the page better (Google Search Central). So rewriting the meta tag alone often changes nothing. Rewrite the title element to match the words people actually searched, and rewrite the first paragraph so it answers the query plainly. The SERP snippet preview shows how both will be cut off.
3. Pages with traffic and no next step
Filter for pages with steady clicks and zero or near zero conversions. Then read each one as a visitor would. Very often the page answers the question well and then ends, with no link to the product, service or signup that the answer naturally leads to.
The fix is one relevant next step, placed after the answer and again at the end. Not a popup, not a banner, a sentence and a link that matches what the reader came for. Conversion rate optimization and the landing page teardown method go further on what that step should look like.
4. Pages that convert but nobody finds
Now reverse the filter: pages with a strong conversion rate and few clicks. These are the most underrated rows in the sheet. Every extra visitor you send them is worth more than a visitor to an average page.
They usually suffer from being buried. Check how many internal links point to them and from where. A converting page that is only reachable from a footer or an old archive page will not rank, however good it is. Link to it from the pages in pattern one and pattern three, where the topic fits.
5. Pages ranking for the wrong question
Open the queries for your highest impression pages and read them. Sometimes a page ranks for searches it does not really answer: a pricing question landing on a feature page, a how to question landing on a product page. The impressions look healthy and the clicks are poor, or the clicks arrive and leave.
Either add a clear section that answers that query on the page, or create the page that should own it and link to it prominently. Keep in mind that the query list is incomplete, because anonymized queries are not shown, so treat it as a sample of intent rather than the full picture (Search Console Help).
6. Two pages splitting one query
When two of your pages appear for the same main query and neither ranks well, they are often competing with each other. Compare them. If one is clearly stronger, fold the useful parts of the weaker one into it and send the old URL to the new one with a permanent redirect, which Google treats as a signal that the target should be shown in results (Google Search Central). If both serve genuinely different intents, sharpen each one so the difference is obvious from the title and the first paragraph.
Step three: decide what to fix first
You will end up with more flagged rows than you can work on. Rank them with two questions.
What is a visitor to this page worth? Use the conversion rate and the value of a conversion from your own data. A page that converts well deserves attention before a page that does not, even if the second one has more traffic.
How big is the realistic gain? A page at position 11 with high impressions has more room than a page at position 4 with few. A page with no next step has more room than one with a weak next step.
Work from the top of that ranking down. The top of that list is where the time goes, which is why the audit should end with a short list, not a long one.
Step four: change one thing and measure it properly
Make one change per page, write down the date, and leave it alone. Changing three things at once tells you nothing about which one worked.
Search Console lets you compare two date ranges directly from the Date filter (Search Console Help). Compare the same number of weeks before and after the change, and look at clicks and conversions rather than position. Position moves for reasons outside your control. Clicks and conversions are what you were trying to improve.
Give each change enough time for the page to be crawled and the results to settle before judging it. If a change did nothing, revert it or try the next idea on the list. If it worked, look for the same pattern on other pages.
Mistakes that waste an audit
Rewriting pages that are working. If a page ranks well and converts, leave it. An audit is not a reason to touch everything.
Treating average position as the goal. It is a diagnostic. Revenue is the goal.
Deleting low traffic pages without checking conversions. Some of the most valuable pages on a site get very little traffic. Pattern four exists for that reason.
Writing new content first. New pages start from zero. Pages already in the results start with the hardest part done, which is why SEO behaves like compound interest and why this audit looks backward before it looks forward.
Running it once. Do it every quarter. The patterns shift as pages climb, competitors publish and search results change.
What you should have at the end
One sheet, one short list of pages, one specific change for each, and a date next to every change. That is an audit someone can act on this week.
If you would like help reading your own data, or want a team to run the whole process, get in touch with me and the X Firm team. For the work that comes after, start with the keyword research framework and content that ranks and converts.