Blog Content Audit Framework: Build a Reliable Url Inventory from Ga4 and Search Console Data

A content audit is only as reliable as the URL inventory beneath it. If your inventory is incomplete, duplicated, outdated or built from one data source alone, decisions about pruning, merging and updating pages can become guesswork. That is particularly risky when you are investigating keyword cannibalisation, duplicate keyword targeting and search intent overlap.

A dependable framework combines Google Analytics 4, Google Search Console, your XML sitemap, crawl data and the CMS export. GA4 helps you understand whether URLs attract engaged users and conversions. Search Console shows how Google displays those URLs for queries, impressions, clicks and positions. Each source tells you something different.

This guide explains how to build a joined, decision-ready inventory for a blog content audit. It also shows where SEOLetters fits into the workflow, particularly when you need to turn audit findings into a structured publishing, internal linking and content-refresh programme.

Why a Reliable URL Inventory Matters in a Content Audit

A URL inventory is the master list of pages that might influence organic performance. It should not simply be a list copied from your sitemap. That approach misses pages with traffic, backlinks, indexed status or historical value, while including URLs that should never be treated as active editorial assets.

A serious inventory should help you answer questions such as:

Which blog URLs currently attract organic clicks?
Which pages receive impressions but fail to win clicks?
Which URLs generate engaged sessions, leads or sales?
Which pages target similar keywords?
Where is search intent overlapping?
Are internal links sending conflicting signals?
Which pages are indexed but have no meaningful performance?
Which URLs should be updated, merged, redirected or left alone?
Which content gaps should become part of the next publishing campaign?

That is the practical purpose of the framework. You are not collecting rows for their own sake. You are building a decision system.

The risk of auditing only one data source

Each platform has gaps:

Data source	What it reveals	What it does not reliably reveal
GA4	Users, sessions, engagement, conversions, landing page behaviour	Exact organic queries and complete historical rankings
Search Console	Queries, impressions, clicks, CTR, average position, indexed page visibility	On-page engagement, revenue quality or every URL on the site
XML sitemap	URLs the site owner wants crawled	Whether those URLs receive traffic or deserve to remain live
Site crawler	Status codes, canonicals, titles, links, technical signals	Real user behaviour and search demand
CMS export	Publication dates, authors, categories, content status	Search visibility and actual business value
Backlink tools	Referring domains and link authority	Whether the page satisfies search intent

The inventory becomes useful when these sources are joined around a canonical URL. That sounds simple. It rarely is.

Trailing slashes, HTTP versions, query parameters, tracking tags, uppercase paths, redirects and canonical tags can split one page into several rows. Fix that before you interpret performance data.

The Data Model: What Your Inventory Should Contain

Before exporting anything, define the fields you need. A good inventory combines technical, editorial, search and business information without becoming so complicated that nobody maintains it.

At minimum, create these column groups:

URL identity fields

Final URL
Normalised URL
URL path
Hostname
Page type
HTTP status code
Canonical URL
Indexability status
Sitemap presence
Redirect target
CMS content ID

GA4 performance fields

Users
Sessions
Engaged sessions
Engagement rate
Average engagement time
Key events or conversions
Revenue, where relevant
Organic landing sessions
Last recorded session date

Search Console fields

Impressions
Clicks
Click-through rate
Average position
Query count
Top query
Query groups
Branded versus non-branded clicks
First and last observed dates

Content and decision fields

Page title
H1
Publication date
Last updated date
Word count
Author or subject expert
Primary topic
Target keyword
Search intent
Content cluster
Internal links in
Internal links out
Backlinks
Recommended action
Priority score
Owner
Review date

This structure may look extensive, but it prevents the usual problem where a team identifies an underperforming URL and then has to open five different tools to understand it.

A practical canonical URL rule

Choose one standard format before joining datasets. For most websites, that means:

https://www.example.com/blog/example-article/

Your normalisation process should usually:

Convert the scheme and hostname to lowercase.
Remove tracking parameters such as utm_source.
Remove unnecessary query strings.
Standardise trailing slashes.
Resolve redirects to the final destination.
Use the canonical URL where the page is valid and self-referencing.
Preserve meaningful parameters only when they represent separate indexable pages.
Flag conflicting canonicals instead of silently overwriting them.

Do not delete every query parameter automatically. Some sites use parameters for legitimate content variations. This is where a crawl and technical review still matter.

Using GA4 to Establish Content Value

GA4 should be used to measure what happens after a user lands on a page. It is not a substitute for Search Console, and its organic landing page data will not tell you which exact query produced each visit.

Still, GA4 is essential because search visibility does not equal business value.

A page may rank well, attract thousands of impressions and still contribute little to the business. Another page might attract modest traffic but produce qualified enquiries every month. Your inventory needs both types of evidence.

GA4 dimensions and metrics to export

For a blog audit, use a report or exploration containing:

Landing page plus query string
Session default channel group
Date
Users
Sessions
Engaged sessions
Engagement rate
Average engagement time per session
Key events
Session conversion rate
Purchase revenue or lead value, if available

Filter for organic traffic when creating the organic landing page view. Keep a second all-channel view if you want to understand whether a URL has value beyond search.

A useful export period is:

Last 12 months for stable performance
Previous 12 months for trend comparison
Last 90 days for recent movement
Last 28 days for urgent changes or algorithm impacts

Longer periods are important for seasonal blogs. A Christmas gift guide should not be judged against a quiet summer month.

GA4 limitations you need to account for

GA4 landing page dimensions can contain query strings and URL variants. This creates duplicate rows unless you remove tracking parameters and align the path with your canonical inventory.

You may also see:

(not set) values
Internal search URLs
Login or account pages
Broken URLs receiving referral traffic
URLs with fragments
Redirecting landing pages
Pages with traffic but no organic sessions

That last category matters. A URL with no organic traffic is not automatically a failure. It could attract direct, referral, email or paid traffic, or it could have valuable assisted conversion activity.

GA4 content value score

You can create a basic score to compare URLs without pretending the metrics are perfectly equal:

Content value score =
(Organic sessions × 0.25)
+ (Engaged sessions × 0.20)
+ (Key events × 0.35)
+ (Revenue or lead value × 0.20)

Normalise each metric first if the numbers vary dramatically. For example, use percentile ranks rather than raw totals when one page has 100,000 sessions and another has 300.

Treat this as a prioritisation device, not an objective truth. A score helps you sort the audit. It does not replace editorial judgement.

Using Search Console to Detect Keyword Cannibalisation

Search Console is the main source for investigating whether multiple URLs appear for the same query set. Google does not label a situation as cannibalisation, though. You need to interpret the page and query relationship.

Export the Performance on Search results report with:

Search type: Web
Date range: ideally 12 months
Dimensions: Page and Query
Metrics: Clicks, impressions, CTR and average position

Search Console has export limits in the interface, and aggregated totals may not equal the sum of all visible query rows. For larger sites, use the Search Console API or a BigQuery export where available.

What potential cannibalisation looks like

A possible cannibalisation pattern usually contains several of these signals:

Two or more URLs receive impressions for the same non-branded query.
The pages have substantially similar search intent.
Rankings alternate between URLs over time.
Neither URL has a stable position despite reasonable impressions.
Clicks are split between pages rather than concentrated.
The weaker page has few unique queries.
Titles and headings target nearly identical terms.
Internal links point to different pages using similar anchor text.
Google’s selected canonical differs from the page you consider primary.

One shared query does not prove cannibalisation. A large website may have several pages ranking for a broad term for valid reasons. The important question is whether the pages serve the same user need and whether the split appears to weaken performance.

Query overlap calculation

For two URLs, calculate the overlap between their query sets:

Query overlap percentage =
(shared queries ÷ smaller URL's total queries) × 100

Example:

URL	Total ranking queries	Shared queries	Overlap
`/blog/seo-audit-guide/`	120	45	37.5%
`/blog/technical-seo-audit/`	80	45	56.3%

The smaller query set is used as the denominator because it shows how much of that page’s visibility is duplicated by the other page.

This calculation is only a starting point. Weight shared queries by impressions or clicks to identify meaningful overlap rather than obscure terms that appear once.

Weighted query overlap

A stronger version gives more importance to queries that generate demand:

Weighted overlap =
shared query impressions ÷ total impressions for the smaller URL

You can also compare:

Shared clicks
Shared impressions
Difference in average position
Difference in CTR
Conversion value by landing page

A URL pair with 70% shared impressions and almost identical intent deserves immediate investigation. Two pages sharing 15 low-volume queries may not.

Building the Combined URL Inventory

The most dependable process is to create one master URL table and append each data source against the normalised URL.

Step 1: Export every possible URL source

Collect:

GA4 landing pages
Search Console pages
XML sitemap URLs
CMS published posts
CMS drafts and archived content, if relevant
Crawl URLs
Redirect destinations
URLs with backlinks
URLs listed in internal links

Do not begin by filtering to URLs with traffic. Pages with no current visibility may still be important because they are new, seasonally dormant, or linked from high-authority pages.

Step 2: Create a URL normalisation column

In a spreadsheet, you can use formulas or Power Query. In a database, use a URL cleaning function.

A conceptual transformation looks like this:

raw_url
→ lowercase hostname
→ remove tracking parameters
→ remove fragments
→ standardise slash format
→ resolve redirects
→ match canonical URL
→ assign url_id

The url_id should be stable. A numeric CMS ID is useful where available, but the normalised canonical URL can work for smaller sites.

Step 3: Union the URL sets

Create a master table containing every unique normalised URL. Then add source flags:

URL	GA4	Search Console	Sitemap	CMS	Crawl	Backlinks
`/blog/a/`	1	1	1	1	1	1
`/blog/b/`	0	1	1	1	1	0
`/blog/c/`	1	0	0	0	1	1

These flags immediately expose gaps. A page in Search Console but not in the sitemap may be an old URL, a forgotten landing page or an indexing issue. A CMS post absent from Search Console may simply be new, or it may have no visibility.

Step 4: Join performance metrics

Use left joins from the master URL set. This is important. If you start from GA4, you will exclude pages with no recorded sessions. If you start from Search Console, you will exclude pages that have no impressions but still exist in the CMS.

Use zeroes for genuinely measured zero values and blanks for unavailable data. Those are not the same:

0 clicks means Search Console recorded the URL but no clicks.
Blank Search Console data may mean the URL was not returned in the export.
0 conversions means there were sessions but no conversions.
Blank GA4 data may mean no session data was available.

That distinction prevents false conclusions.

The Recommended Inventory Schema

The following structure works well for a content audit spreadsheet, database or data warehouse.

Field	Example	Why it matters
URL ID	4832	Stable record matching
Canonical URL	`https://site.com/blog/a/`	Primary page identity
Page type	Blog article	Scope and segmentation
Status	200	Technical eligibility
Indexable	Yes	Search availability
Canonical match	Self	Canonical confidence
Sitemap	Yes	Crawl management
GA4 organic sessions	1,240	Organic usage
GA4 engaged sessions	930	Quality of visits
Key events	42	Business outcome
GSC clicks	860	Search demand captured
GSC impressions	24,500	Search visibility
GSC CTR	3.5%	SERP appeal
GSC average position	8.7	Ranking strength
Top query	content audit framework	Primary visibility
Query count	186	Breadth
Intent	Informational	Page purpose
Topic cluster	Content audits	Authority structure
Last updated	2023-04-18	Freshness review
Inbound internal links	7	Site prominence
Decision	Refresh	Action
Priority	High	Workflow order

Do not overload this table with every possible content note. Keep a separate decision log if detailed recommendations become difficult to scan.

A Cannibalisation Audit Checklist

Use this checklist for every cluster of potentially overlapping pages:

Do the URLs rank for the same core query?
Do they satisfy the same search intent?
Are the audiences materially different?
Is one page a guide and the other a narrow supporting article?
Does one URL have substantially stronger backlinks?
Does one page attract most of the clicks?
Are impressions divided across both pages?
Have rankings alternated between the URLs?
Do title tags promise almost the same answer?
Do H1s and subheadings target duplicate keyword targeting?
Do internal links use similar anchor text to different destinations?
Is there a canonical conflict?
Does either page have unique conversions or backlinks?
Could the pages be merged without losing useful intent coverage?
Would a redirect remove a valuable standalone asset?

The checklist is deliberately cautious. Merging pages because their keywords look similar can remove useful coverage. Keyword similarity is an indicator, not a verdict.

Classifying Search Intent Overlap

Search intent overlap is best evaluated by looking at the actual result pages, not only keyword tools. Search for the overlapping query and inspect what Google is rewarding.

Classify each URL using four practical intent categories:

Intent category	Typical user need	Suitable page format
Informational	Learn a concept or process	Guide, explainer, tutorial
Commercial investigation	Compare options or assess solutions	Comparison, review, buyer guide
Transactional	Take an action or buy	Product, service or category page
Navigational	Find a known brand or resource	Brand or destination page

Two pages may both include the word “audit” while serving different needs. One could explain how to conduct a technical audit. Another could compare audit software. Their wording overlaps, but the desired result is different.

Search intent scoring rubric

Score each URL pair from 0 to 3:

Criterion	0	1	2	3
Query overlap	None	Low	Moderate	High
Intent similarity	Different	Partly related	Mostly similar	Identical
Topic similarity	Different	Adjacent	Closely related	Same subject
Ranking volatility	Stable	Mild	Noticeable	Frequent switching
Internal linking conflict	None	Minor	Several conflicts	Strongly conflicting
Conversion duplication	None	Unclear	Some overlap	Same outcome

A total of 12 or more suggests a high-priority investigation. Scores between 7 and 11 need a manual review. Lower scores usually indicate related content rather than true cannibalisation.

Do not treat the score as an algorithmic answer. It is a way to make reviews consistent across a large blog.

Internal Linking Conflicts and Ranking Signal Dilution

Internal linking conflicts are easy to miss because every individual link may appear reasonable. The problem emerges at cluster level.

Suppose five articles use the anchor text “content audit framework”. Three point to a broad guide, while two point to a narrower article. Search engines receive a less consistent indication of which page should act as the primary resource. Users may also land on a page that does not match the anchor’s promise.

This can contribute to ranking signal dilution, particularly when:

Several pages target the same primary topic.
Link equity is distributed across thin or outdated URLs.
Important pages receive fewer internal links than weaker alternatives.
Breadcrumbs, related posts and in-content links point in competing directions.
Anchor text implies one topic while the destination covers another.

Internal linking review process

Export all internal links from a crawl.
Group links by source URL and destination URL.
Extract anchor text and surrounding context.
Identify destination pages for each important topic.
Compare the link pattern with your intended content hierarchy.
Redirect or consolidate outdated pages where appropriate.
Update contextual links after content decisions are approved.
Re-crawl and verify the final structure.

A content audit decision is incomplete until internal links support it. Merging two articles without updating internal links leaves the old structure in place and can create redirect chains or confusing references.

Turning Inventory Findings into Content Decisions

Once the inventory is complete, assign one clear action to every in-scope URL. Avoid vague labels such as “improve” unless you also define what improvement means.

Decision	Use when	Typical implementation
Keep	Page performs well and intent is clear	Monitor and maintain
Refresh	Page has demand but weak quality or outdated information	Update facts, examples, structure and links
Expand	Page ranks for related terms but lacks depth	Add useful sections based on query data
Consolidate	Several URLs overlap heavily	Merge strongest assets into one primary page
Redirect	A redundant URL has little unique value	301 redirect to the most relevant page
Repurpose	Intent has changed or format is wrong	Rebuild as comparison, landing page or guide
Noindex	Useful to users but unsuitable for search	Apply carefully, not as a shortcut
Remove	No value, demand, links or business purpose	Return an appropriate status code

The strongest-page rule for consolidation

When consolidating, choose the primary page using evidence:

Higher organic clicks
Better conversion quality
Stronger backlink profile
More relevant search intent
Better historical performance
More accurate URL and title
Greater topical completeness
Stronger internal link position

Traffic alone should not decide. A page with fewer visits but much higher lead quality may be the better destination.

Example: two competing blog articles

Imagine a software company has:

/blog/content-audit-guide/
/blog/how-to-audit-blog-content/

The first receives 1,900 organic clicks and ranks for 210 queries. The second receives 340 clicks and ranks for 78 queries. Both target informational intent, use similar headings and receive impressions for “content audit”, “blog content audit” and “content audit checklist”.

A sensible recommendation might be:

Retain the first URL as the primary guide.
Move unique examples and useful sections from the second article into it.
Preserve any queries with a distinct sub-intent.
Redirect the second URL after reviewing backlinks and conversions.
Update internal links to point consistently to the primary guide.
Monitor clicks, impressions and ranking distribution for 8 to 12 weeks.

That is not simply deleting an article. It is restructuring the cluster.

How SEOLetters Supports the Audit-to-Publishing Workflow

A URL inventory tells you what needs attention. The next challenge is executing the work consistently across a large content operation.

SEOLetters, the AI blog writer for structured SEO publishing can help you move from keyword research and content decisions to publishable articles, internal links, images, schema and direct deployment. You can bring your own AI keys and route different stages to Gemini, OpenAI or Claude, which gives your team more control over the workflow.

This matters after a cannibalisation audit because the required work is usually not one article. You may need to:

Rewrite a primary guide.
Produce supporting cluster content.
Create content briefs from query gaps.
Refresh several declining URLs.
Add product-aware sections.
Insert internal links based on the revised hierarchy.
Publish to WordPress, Shopify or a webhook.
Maintain a scheduled content-refresh campaign.

SEOLetters is designed around that full sequence rather than isolated text generation. Its keyword research includes difficulty ratings, while topical authority clusters help turn a single subject into a structured plan.

Use the audit to create a content campaign

After assigning decisions, group work into campaigns:

Campaign type	Inputs from the audit	Output
Consolidation campaign	Overlapping URLs and primary destination	Merged article and redirects
Refresh campaign	Declining traffic, old data, weak CTR	Updated existing pages
Authority campaign	Topic gaps and supporting queries	Cluster of new articles
Internal link campaign	Conflicting destinations	Revised site architecture
Commercial campaign	Informational pages with product relevance	Product-aware supporting content

The autonomous campaign scheduler in SEOLetters for recurring blog content and refresh campaigns can be configured around a topic, cadence and publishing destination. It can research, write and publish on schedule, while refresh campaigns help keep existing pages current.

That is useful for teams that have completed an audit but lack the operational capacity to update 50 or 100 pages manually. The strategy still comes from you. The repeatable production work is handled inside the publishing workflow.

Quality Controls Before Publishing Audit Recommendations

Automation can accelerate execution, but it does not remove the need for editorial and SEO checks. Every merged or refreshed page should pass a quality review.

Content quality checks

Does the article answer the dominant search intent?
Are claims supported by credible sources?
Are statistics dated and correctly attributed?
Is the advice specific enough to act on?
Does the introduction accurately describe the outcome?
Are examples relevant to the audience?
Are headings structured logically?
Are competing pages addressed or removed?
Does the page demonstrate first-hand experience where appropriate?
Is the content genuinely better than the pages it replaces?

Technical checks

Is the final URL returning a 200 status?
Does the page use a self-referencing canonical?
Is the redirect direct rather than chained?
Has the XML sitemap been updated?
Are old internal links corrected?
Are images compressed and given useful alt text?
Is Article or relevant schema valid?
Does the page work on mobile?
Are there duplicate title tags or H1s?
Is the page linked from the relevant cluster hub?

Measurement checks

Record a baseline before changes:

Organic clicks
Impressions
CTR
Average position
Organic sessions
Engagement rate
Key events
Assisted conversions
Ranking query count
Number of URLs appearing for the target query

Without a baseline, a later performance change is difficult to interpret. Seasonality and algorithm updates can distort the result, so annotate the implementation date and compare an appropriate period.

Measuring Whether Cannibalisation Has Improved

Do not judge consolidation after a few days. Search systems need time to recrawl, reassess and redistribute visibility.

A practical review schedule is:

Week 1 to 2: Check redirects, indexing, canonical selection and crawl errors.
Week 3 to 6: Review impressions, query coverage and early ranking movement.
Week 7 to 12: Compare clicks, CTR, conversions and ranking stability.
After 3 months: Decide whether further content, links or intent adjustments are needed.

Track the query group rather than only the surviving URL. The key question is whether the topic now has stronger visibility and better outcomes.

Cannibalisation monitoring dashboard

KPI	Before consolidation	After consolidation	Interpretation
Combined clicks	1,240	1,780	Topic visibility increased
Combined impressions	34,000	41,500	Broader exposure
Weighted CTR	3.6%	4.3%	Better SERP alignment
Average position	14.2	9.8	Stronger ranking
Ranking URLs for target term	3	1	Less internal competition
Key events	18	31	Better business value
Internal links to primary page	9	22	Clearer site signal

This type of dashboard gives stakeholders something more useful than a statement that “the content was improved”.

Common Errors in GA4 and Search Console Audits

Treating average position as an exact rank

Search Console average position is aggregated across queries, devices, locations and impressions. A page with an average position of 7 may rank first for one term and 30th for another.

Use query-level data and group terms by intent. A single average can conceal the issue you are trying to diagnose.

Assuming two pages with the same keyword are cannibalising

Modern pages rank for many related terms. Duplicate keyword targeting is not automatically a problem, particularly when two URLs serve different audiences or stages of the buying journey.

Review the result pages and the page purpose before consolidating.

Using GA4 traffic as the only quality signal

A page with low engagement may be answering a simple question efficiently. A page with high engagement may still fail to generate leads. Consider scroll behaviour, next-page visits, key events and assisted conversions together.

Ignoring URLs with no data

New content, seasonal pages and pages with poor indexing can appear empty in GA4 and Search Console. Keep them in the inventory and mark the relevant fields as unavailable or zero based on what the platform actually reports.

Forgetting historical URLs

Old URLs can have valuable backlinks and residual ranking signals. A URL absent from the current sitemap may still deserve a redirect, refresh or consolidation review.

Publishing new articles before resolving overlap

Adding another article to a confused cluster often increases the problem. Run the inventory and intent review first, then decide whether the next asset should be new content or a stronger version of what already exists.

A Repeatable Quarterly Content Audit Framework

You can run this process every quarter or twice each year, depending on site size and publishing velocity.

Phase 1: Assemble

Export GA4 landing page data.
Export Search Console page and query data.
Crawl the site.
Export the sitemap.
Pull CMS and backlink data.
Create the normalised URL master list.

Phase 2: Diagnose

Segment blog URLs from product, category and utility pages.
Identify declining, emerging and high-value pages.
Group URLs by topic and intent.
Calculate query overlap.
Review internal linking conflicts.
Flag technical inconsistencies.

Phase 3: Decide

Assign one action to each URL.
Select primary pages for overlapping clusters.
Record redirect destinations.
Define refresh requirements.
Identify content gaps.
Prioritise by business value, effort and search opportunity.

Phase 4: Execute

Rewrite or expand priority pages.
Merge and redirect confirmed duplicates.
Update titles, headings and metadata.
Repair internal links.
Add relevant schema and images.
Publish new cluster content where a genuine gap exists.

Phase 5: Measure

Check indexing and technical implementation.
Compare query groups over time.
Track clicks, impressions, CTR and conversions.
Review ranking URL count.
Re-score the cluster after 8 to 12 weeks.
Add findings to the next campaign.

For teams publishing internationally, SEOLetters for multi-language SEO content operations can support generation across 21 languages while keeping campaigns connected to topics, destinations and publishing schedules. That is particularly relevant when the same cannibalisation issue appears across several country folders or translated versions.

Final Takeaway: Treat the URL Inventory as a Decision Engine

A dependable blog content audit starts with data integrity. GA4 shows how visitors behave and whether pages contribute to business outcomes. Search Console reveals query visibility and possible page competition. Crawls, sitemaps, CMS records and backlink data complete the picture.

The main lessons are straightforward:

Build from every relevant URL source.
Normalise URLs before joining data.
Keep unavailable data distinct from zero performance.
Investigate intent, not just keyword similarity.
Measure query overlap using impressions and clicks.
Review internal linking conflicts before creating more content.
Consolidate only when the evidence supports it.
Track the topic after implementation, not only the surviving URL.
Use scheduled refreshes to prevent the inventory becoming stale.

If you are managing a growing blog, the audit itself is only one part of the work. The harder requirement is turning decisions into a disciplined publishing operation that produces, updates and distributes content consistently.

You can use SEOLetters to support that next stage, from keyword research and topical authority planning to article generation, internal links, schema, images, multilingual production and direct publishing. If you need a larger workflow review, use the rightbar as the contact path and connect the audit findings to a measurable content campaign.