Ai Overviews and Organic Traffic Attribution: What Search Console and Analytics Data Cannot Reveal

Google’s AI Overviews are changing the relationship between search visibility, clicks and organic traffic measurement. A page can contribute to an answer, influence a later search, support a brand decision or help a user compare products without generating a clean, attributable visit in Search Console or Google Analytics.

That creates a difficult reporting problem. Your dashboards may show falling clicks for a query, while your brand is appearing in an AI-generated answer. Several pages may also be eligible for the same topic, making it unclear which URL deserves credit. This is where keyword overlap SEO, duplicate keyword targeting and keyword cannibalisation become much harder to diagnose.

You need a measurement framework that accepts the limits of the data. Search Console and Analytics remain essential, but neither system can reveal the full path from AI Overview exposure to eventual conversion. They report observable events. They do not report every influence that happened before the click.

This guide explains what AI Overviews can obscure, how to interpret organic data safely, how keyword cannibalisation affects attribution, and how to build a more reliable workflow with content planning and publishing tools such as SEO Letters.

Why AI Overviews Make Organic Traffic Attribution More Difficult

AI Overviews place a generated summary directly within the search results page. The summary may answer part of the query, include several supporting links and encourage the user to continue researching without visiting any cited page immediately.

That changes the traditional SEO funnel:

1. A user searches for a topic.
2. Google displays an organic result.
3. The user clicks the result.
4. Analytics records a session.
5. Search Console records an impression and click.
6. The user completes a conversion.

The AI Overview introduces additional stages:

1. A user searches for a topic.
2. Google generates an AI Overview.
3. Your page may be used as a supporting source.
4. The user reads the answer without clicking.
5. The user searches your brand or a related question later.
6. The user eventually visits through another channel.
7. Analytics may attribute that visit to direct, organic brand, referral, paid or another source.

The first influence may be invisible. The later visit may be visible but incorrectly interpreted as an independent interaction.

This is the central measurement issue. Organic traffic attribution records identifiable visits, not every form of search influence.

What Search Console Can Reveal About AI Overview Performance

Search Console remains useful because it provides query, page, country, device, impression, click, CTR and average-position data. It can help you identify changes in search demand and performance across pages.

However, it does not currently provide a complete, separate reporting layer for every AI Overview appearance or influence. Depending on the reporting environment and feature behaviour, AI Overview activity may be blended into broader search reporting rather than exposed as a clean, universally available dataset.

Search Console can usually help you analyse:

• Search queries associated with your pages.
• Impressions and clicks from Google Search.
• URL-level performance.
• Changes in CTR and average position.
• Country and device patterns.
• Search demand for branded and non-branded terms.
• Pages that gain or lose visibility around a topic.
• Query groups where impressions rise but clicks fall.

It cannot reliably tell you:

• Whether your URL was cited inside an AI Overview.
• Whether a user read your brand or advice in an AI-generated summary.
• Whether a later branded search was caused by an earlier AI Overview exposure.
• Whether your content was considered during answer generation but not cited.
• Which cited source contributed the most to a user’s eventual conversion.
• Whether the user had already formed a preference before clicking your listing.
• How many users saw your content indirectly through an AI-generated synthesis.

That distinction matters when you interpret falling CTR.

A lower CTR may mean the result became less attractive. It may also suggest that the AI Overview answered enough of the question that fewer users needed to click. Those are different commercial situations, and the available data may not distinguish them.

What Google Analytics Can and Cannot Attribute

Google Analytics 4 is designed to report interactions on your site. It can track sessions, landing pages, engagement, events, conversions and user journeys when the user reaches your website and the tracking implementation works correctly.

It cannot track a user who only sees an AI Overview and leaves the results page.

This creates several blind spots:

Activity	Likely visibility in Search Console	Likely visibility in Analytics
AI Overview appears without your site being cited	Limited or unavailable	None
Your page is cited but not clicked	May be unclear or blended	None
User clicks your citation	Potentially visible as a search interaction	Often appears as organic traffic
User sees your citation, later searches your brand and visits	Search data may show later query	May appear as organic brand or direct traffic
User clicks, returns later and converts	Initial click may be visible	Conversion path may be partially visible
User discusses your brand after seeing an AI answer	None	None unless they visit
User visits through an untagged third-party link after AI exposure	None	May be direct or referral traffic

Analytics can show what happened after a site visit. It is much less capable of showing why the visit happened.

That is not a flaw in the platform. It is a limitation of observable data. When it comes to AI-assisted search, teams need to stop treating a tracked click as the beginning of influence and a conversion as proof of a single-channel cause.

The Difference Between Visibility, Clicks and Influence

SEO reporting often groups several separate concepts under the word “performance”. That makes the reporting look simpler, but it also causes bad decisions.

You should separate at least four layers:

1. Search visibility

This is your presence in a search environment. It can include:

• Traditional organic listings.
• Featured snippets.
• People Also Ask results.
• Video or image results.
• AI Overview citations.
• Brand mentions within generated answers.
• Entity-level visibility for your company or product.

Some of this visibility is measurable. Some is only partially observable.

2. Search interaction

This refers to an action taken in the search interface, such as:

• Clicking an organic result.
• Expanding a result.
• Selecting a cited source.
• Refining the query.
• Searching your brand after reading an answer.

Only some of these interactions are available to your reporting stack.

3. Website engagement

This begins when a user reaches your website. It may include:

• Engaged sessions.
• Scroll depth.
• Product views.
• Form submissions.
• Downloads.
• Account creation.
• Assisted conversions.
• Returning visits.

4. Commercial influence

This is the broader effect of content on consideration and choice. It may occur when a user:

• Learns your terminology.
• Recognises your brand later.
• Uses your comparison framework.
• Returns to a product page directly.
• Mentions your business to a colleague.
• Converts after several untracked interactions.

The last category is usually the least visible, yet it can be commercially important. A sensible SEO dashboard should not pretend that all four layers can be measured with the same precision.

How AI Overviews Interact With Keyword Cannibalisation

Keyword cannibalisation occurs when multiple pages on the same website compete for similar queries or overlapping search intents. In a traditional search result, this can lead to URL switching, diluted relevance signals, unstable rankings and internal competition.

AI Overviews add another layer. Google may draw supporting information from several pages on your site, even when those pages target closely related queries. Your content could be contributing to the answer as a group, while your own reports show traffic moving between URLs.

This can look like a ranking problem when it is actually a content architecture problem.

For example, imagine a software company with these pages:

• /ai-content-writing-guide/
• /best-ai-blog-writer/
• /ai-seo-writing-tools/
• /automated-content-publishing/
• /ai-content-generation-software/

All five URLs may mention AI writing, SEO articles, automation and publishing. If the pages do not have clear roles, Google may alternate between them for overlapping queries. One page may receive the visible click, while another supports the topical context or appears in an AI Overview citation.

Your analytics data will usually credit the page that received the visit. It will not show the combined contribution of the other pages.

Keyword overlap SEO is not always a problem

Some overlap is natural. A well-organised topic cluster will contain related pages that discuss connected concepts.

The problem appears when two pages have:

• The same primary intent.
• Similar titles and headings.
• Closely matching target terms.
• Identical conversion goals.
• Overlapping internal links.
• Similar content depth.
• No clear parent and child relationship.

A page about “AI blog writing software” can coexist with a page about “automated SEO content publishing” if the user needs are different. It becomes a cannibalisation risk when both pages attempt to rank for the same product-comparison intent and lead to the same conversion path.

Why AI Overviews Can Hide Cannibalisation Signals

Traditional cannibalisation audits usually compare impressions, clicks, average position and URL changes for a query. That remains useful, but AI Overviews can make the evidence less conclusive.

Consider this pattern:

Month	Page A clicks	Page B clicks	Combined impressions	Combined CTR
January	420	160	18,000	3.2%
February	310	245	20,500	2.7%
March	270	190	19,800	2.3%

A basic reading might conclude that Page A lost rankings and Page B gained them. A deeper reading asks:

• Did AI Overview coverage increase for this topic?
• Did both URLs appear for similar queries?
• Did the search intent change?
• Did users get enough information without clicking?
• Did branded searches rise after non-branded impressions?
• Did the pages split authority because their content roles were unclear?
• Did Google consolidate the result into one preferred URL while drawing context from both?

Search Console cannot answer every question directly. You need to combine query analysis, page-level content review, SERP observation, internal-link analysis and conversion data.

A Practical Keyword Cannibalisation Audit for AI Search

A useful audit should not focus only on rankings. It should assess intent, page purpose, search demand, content similarity and business outcomes.

Step 1: Export query and page data

Use Search Console to export at least the previous 16 months of data where available. Include:

• Query.
• Page.
• Impressions.
• Clicks.
• CTR.
• Position.
• Country.
• Device.
• Search appearance where available.

Then group queries by topic rather than reviewing them as isolated phrases.

Step 2: Identify duplicate keyword targeting

Look for queries where multiple URLs receive meaningful impressions. Set a working threshold so that you do not waste time on trivial impressions.

A simple review threshold might include:

• At least 100 impressions for the query and period.
• At least two URLs receiving impressions.
• At least one URL receiving clicks.
• A clear commercial or strategic relationship to the topic.

The exact threshold should reflect your site size. A large publisher may need a higher minimum. A small business site may need to inspect every meaningful query.

Step 3: Map search intent

Classify each query according to the user’s likely need:

Intent category	Typical query language	Suitable content format
Informational	what is, how does, guide, examples	Guide, explainer or glossary
Commercial investigation	best, alternatives, comparison, review	Comparison or evaluation page
Transactional	buy, pricing, software, service	Product or service page
Navigational	brand or product name	Brand, product or login page
Troubleshooting	fix, audit, diagnose, why	Support article or diagnostic guide

This stage is important because identical words do not always represent identical intent. “AI content tool” might indicate product research. “How to use AI for content” is more likely educational.

Step 4: Score URL roles

Assign each page a clear role:

• Primary topic page.
• Supporting cluster page.
• Commercial landing page.
• Comparison page.
• Product page.
• Glossary or definition page.
• Refresh target.
• Consolidation candidate.

If two pages have the same role and the same intent, they deserve close scrutiny.

Step 5: Compare content similarity

Review more than the percentage of matching words. Assess:

• Main question answered.
• Audience.
• Funnel stage.
• Recommended action.
• Unique evidence.
• Internal links.
• Title and H1.
• Structured data.
• Backlink profile.
• Freshness.
• Conversion path.

Two pages can use different wording yet still compete directly. This whole thing is about purpose, not just text duplication.

Step 6: Make a decision

For each overlap cluster, choose one of these actions:

Situation	Recommended action
Pages answer the same intent and one is clearly stronger	Redirect or consolidate into the stronger page
Pages have different intents but similar terminology	Keep both, strengthen differentiation
One page supports another broader page	Rebuild as a clear topic cluster
Both pages have valuable links and distinct audiences	Keep both, improve canonical and internal-link signals
One page is thin and adds little value	Merge, redirect or remove
A commercial page is being outranked by an informational page	Clarify intent, links, copy and page architecture
URLs switch because query meanings vary	Build more precise subtopics rather than forcing one page

Content Consolidation Strategy for AI Overview Visibility

A content consolidation strategy should not be based on the assumption that fewer pages always perform better. The aim is to create clearer relevance signals and a better user journey.

Consolidation may involve:

• Merging two overlapping articles.
• Redirecting outdated URLs.
• Moving a section into a stronger pillar page.
• Rewriting titles to separate intent.
• Creating a comparison page from competing review articles.
• Updating internal links so the preferred URL receives contextual support.
• Removing duplicated FAQs.
• Adding original data, examples and expert commentary to the surviving page.

A consolidation decision should consider backlinks and historical traffic. Do not delete a page simply because its current clicks are low. It may hold useful links, rank for valuable long-tail queries or contribute to the wider topical authority of the site.

A consolidation scoring rubric

Score each page from 1 to 5:

Criterion	Question
Intent clarity	Does the page have a distinct reason to exist?
Organic performance	Does it receive valuable impressions or clicks?
Conversion value	Does it support leads, sales or assisted conversions?
Link equity	Does it have relevant external links?
Content quality	Does it contain unique, accurate and useful material?
AI search relevance	Does it provide concise, source-worthy evidence?
Maintenance cost	Can your team keep it current?

A low score across intent clarity, content quality and conversion value makes consolidation more attractive. A page with strong links and useful evidence may need redevelopment rather than deletion.

What AI Overview Citations Mean for Content Strategy

A citation inside an AI Overview is not equivalent to a conventional position-one ranking. It may produce visibility without a click, and it may expose your brand to users who would not have selected a standard result.

The page still needs to satisfy ordinary SEO requirements. AI-generated answers typically rely on content that is understandable, focused and supported by credible information.

Pages that are easier to interpret often contain:

• A direct answer near the beginning.
• Clear heading hierarchy.
• Definitions in plain language.
• Specific examples.
• Verifiable claims.
• Updated statistics with sources.
• Expert experience or first-hand observations.
• Distinctive frameworks.
• Concise summaries.
• Strong internal links to related explanations.
• Structured product or service information where appropriate.

This does not mean writing only for extraction. You need depth as well as clarity. A page with one neat definition and no useful evidence is unlikely to be a strong resource in its own right.

Building a Better Measurement Model

Your measurement model should distinguish between what you know, what you infer and what you cannot observe.

Known data

This includes:

• Search Console clicks.
• Search Console impressions.
• Landing pages.
• Organic sessions.
• Engaged sessions.
• Assisted conversions.
• Brand and non-brand query trends.
• Revenue from tracked journeys.
• Returning-user behaviour.

Inferred signals

These may suggest AI Overview influence:

• Impressions rising while CTR falls.
• Non-brand visibility increasing while brand searches grow later.
• Organic traffic declining but direct traffic increasing.
• More users entering through deeper commercial pages.
• Query wording becoming more specific.
• Referral traffic from platforms where AI-assisted research may occur.
• Conversion rates improving despite fewer top-of-funnel clicks.
• One topic cluster gaining visibility while individual URLs fluctuate.

These are useful signals, but they are not proof of causation.

Unobservable activity

You may not be able to see:

• An unclicked citation.
• A user reading a generated answer.
• A user remembering your brand.
• A conversation influenced by that exposure.
• A later offline purchase.
• A private search or browser interaction.
• A user copying an answer into another system.

Your reporting should say this plainly. It makes your conclusions more credible.

A Measurement Framework for AI Overviews and Organic Traffic

Use a layered framework rather than one headline traffic number.

Layer 1: Visibility indicators

Track:

• Topic-level impressions.
• Query coverage.
• Brand and non-brand impressions.
• SERP feature presence.
• Estimated AI Overview observations from manual checks or approved monitoring tools.
• Number of pages appearing for a topic.
• Share of queries with a preferred URL.

Layer 2: Click indicators

Track:

• Organic clicks.
• CTR by query class.
• CTR by device.
• CTR by page type.
• Clicks to commercial pages.
• Clicks from informational pages to product pages.
• New versus returning organic users.

Layer 3: Engagement indicators

Track:

• Engaged sessions.
• Time-adjusted engagement.
• Scroll or interaction events.
• Internal-link clicks.
• Product views.
• Form starts.
• Downloads.
• Email sign-ups.
• Return visits.

Layer 4: Commercial indicators

Track:

• Organic leads.
• Sales-qualified leads.
• Revenue.
• Assisted conversions.
• Conversion rate by landing page.
• Conversion rate by content cluster.
• Pipeline influenced by organic content.
• Customer acquisition cost where applicable.
• Revenue per organic session.

Layer 5: Strategic indicators

Track:

• Cannibalisation incidents.
• Number of consolidation candidates.
• Content refresh completion.
• Topic coverage.
• Content decay.
• Internal-link health.
• Percentage of pages with a defined search intent.
• Percentage of articles with a current owner and review date.

This framework prevents a fall in clicks from being treated as an automatic failure. Sometimes the site is receiving less traffic but generating more qualified demand. Sometimes it is losing both visibility and commercial value. The dashboard needs to separate those cases.

How to Use Analytics Without Overclaiming Attribution

GA4 attribution models can distribute credit across tracked touchpoints, but they cannot assign credit to an interaction that never reached your site.

A user may:

1. See your brand in an AI Overview.
2. Search your company name two days later.
3. Click a paid advert.
4. Return through a bookmarked page.
5. Convert through a direct session.

Analytics may show paid, direct or another tracked source depending on the journey and attribution settings. The original AI Overview exposure will probably not appear.

To reduce confusion:

• Separate first user source from session source.
• Review both data-driven and last-click views.
• Compare assisted conversions with final-touch conversions.
• Analyse brand and non-brand organic traffic separately.
• Create content-group reports for topic clusters.
• Monitor conversion lag.
• Review new-user and returning-user patterns.
• Use consistent campaign tagging for email, partnerships and social distribution.
• Annotate major Google updates and significant AI search changes.

Do not use a single attribution model as a factual record of influence. It is a reporting model based on available events.

A Hypothetical Example: Falling Clicks, Rising Commercial Value

A B2B software company publishes a detailed guide about automated content workflows. Before AI Overviews became common for the topic, the article generated:

• 1,800 monthly impressions.
• 240 organic clicks.
• 14 demo requests.
• 4 sales-qualified leads.

Six months later, the page generates:

• 3,600 monthly impressions.
• 190 organic clicks.
• 18 demo requests.
• 7 sales-qualified leads.

The CTR has declined. A traffic-only report would flag the article as deteriorating.

A broader review suggests:

• The page is visible for more high-intent queries.
• Users who do click are more qualified.
• A related product page receives more assisted conversions.
• Branded searches for the company and category have increased.
• Two overlapping articles were recently consolidated.
• The guide is being cited in AI Overview results during manual checks.

The evidence does not prove that AI Overviews caused the additional leads. It does suggest that the relationship between impressions, clicks and commercial value has changed. The correct action may be to improve the conversion path and strengthen the article’s evidence, rather than chase the previous CTR.

How SEO Letters Supports This Workflow

The measurement problem usually begins with a content planning problem. If your site has dozens of articles targeting the same topic, attribution becomes unclear before the data reaches your dashboard.

SEO Letters helps you move from keyword research to structured publishing with keyword difficulty ratings, topical authority clusters and content plans designed around related search intent. That gives you a clearer basis for deciding which page should own a topic and which pages should support it.

The platform can also help you:

• Identify content gaps against competitors.
• Organise related keywords into topic clusters.
• Generate structured articles with headings and internal links.
• Create product-aware content for affiliate and ecommerce publishing.
• Publish directly to WordPress, Shopify or webhooks.
• Schedule recurring campaigns.
• Refresh existing content instead of producing endless new URLs.
• Generate content across 21 languages.
• Track performance through a publishing dashboard.
• Route different workflow stages to Gemini, OpenAI or Claude using your own keys.

This matters because a sensible search intent mapping SEO process should continue after publication. You need to know whether a new article has created useful coverage or just added another page to an already crowded cluster.

Use SEO Letters as an AI blog writer and content consolidation engine

A strong workflow looks like this:

1. Research the topic and related queries.
2. Group terms by intent.
3. Select a primary URL for each major intent.
4. Define supporting pages and their boundaries.
5. Produce the article with a clear heading structure.
6. Add contextual internal links.
7. Publish to the correct destination.
8. Monitor performance and query overlap.
9. Refresh, merge or reposition pages when the data suggests a problem.

The advantage is operational consistency. Your team can apply the same logic to every cluster rather than relying on a spreadsheet that no one updates after publication.

Internal Links, Canonicals and Preferred URL Signals

Internal links will not solve every cannibalisation issue, but they help communicate site structure.

Use internal links to:

• Point broad informational pages towards the relevant commercial page.
• Link supporting articles to the topic pillar.
• Use descriptive anchor text without forcing exact-match phrases everywhere.
• Reduce links between pages that serve the same intent.
• Promote the preferred URL from high-authority pages.
• Connect refreshed content to current product or service pages.
• Build a logical route from discovery to evaluation.

Canonical tags are also useful, but they should not be used as a substitute for content decisions. If two pages are genuinely different and valuable, canonicalising one to the other may remove useful search signals. If they are effectively duplicates, a redirect or consolidation may provide a clearer outcome.

Check:

• Canonical tags.
• Indexability.
• Redirect chains.
• XML sitemaps.
• Internal links.
• Breadcrumbs.
• Hreflang where relevant.
• Structured data.
• Orphaned URLs.
• Parameter and faceted navigation behaviour.

AI search makes clear site architecture more important because multiple pages may be evaluated as a related evidence set.

How to Monitor AI Overview Effects Without Complete Data

You cannot create a perfect AI Overview attribution report with data that the platforms do not expose. You can build a useful monitoring programme.

Run a recurring SERP observation process

For a representative set of queries, record:

• Whether an AI Overview appears.
• Whether your brand is mentioned.
• Whether your page is cited.
• Which competitors are cited.
• Whether citations change by country, device or query wording.
• Whether the overview answers the query fully.
• Whether standard organic results remain visible above the fold.
• Whether the query leads to a follow-up search.

Manual checks are not a replacement for a scalable dataset. They are still valuable for understanding the search environment.

Use a query cohort model

Create groups such as:

• Informational non-brand queries.
• Commercial non-brand queries.
• Brand queries.
• Comparison queries.
• Product-specific queries.
• Existing-customer queries.
• Queries with likely AI Overview intent.
• Queries where your pages overlap.

Compare these cohorts month over month. This avoids drawing conclusions from one volatile keyword.

Add annotations to your reporting

Mark:

• Major content consolidations.
• New topic clusters.
• Algorithm updates.
• Changes to page templates.
• New schema deployment.
• Internal-linking projects.
• Significant AI search feature changes.
• Product launches.
• PR campaigns.
• Paid media changes.

Without annotations, a chart may show a change but not its likely operational cause.

Metrics That Matter More Than Raw Organic Sessions

Organic sessions remain important, especially for publishers and ecommerce sites. They should not be treated as the only measure of SEO health.

Consider adding:

Metric	Why it matters
Non-brand impression growth	Indicates wider topic visibility
High-intent query coverage	Shows whether visibility is commercially useful
Organic assisted conversions	Captures content that supports later action
Revenue per organic session	Connects traffic to value
Engaged organic sessions	Filters out low-quality visits
Internal-link progression	Shows whether informational content supports evaluation
Topic cluster visibility	Reduces dependence on one URL
Cannibalisation rate	Shows how often multiple pages compete
Content refresh lift	Measures the value of updating existing assets
Branded search growth	May indicate wider recognition, though it is not proof of AI influence

Set benchmarks by page type. A glossary page should not be judged by the same conversion target as a pricing page. An informational guide may be successful when it introduces a qualified user who converts later through another page.

Common Reporting Mistakes

Mistake 1: Treating CTR decline as proof of content failure

CTR can fall because of:

• AI Overviews.
• More SERP features.
• Changes in query mix.
• Mobile layout changes.
• Stronger competition.
• A less compelling title.
• Lower ranking.
• Satisfied users who do not need to click.

Diagnose the cause before rewriting the entire page.

Mistake 2: Giving all credit to the final landing page

A product page may receive the conversion, while an earlier guide shaped the user’s decision. Review assisted conversions and path exploration.

Mistake 3: Calling every URL switch cannibalisation

Google may select different pages for different meanings of a broad query. Confirm intent overlap before merging anything.

Mistake 4: Publishing more pages to recover lost clicks

This often increases duplicate keyword targeting. The site becomes larger, but the topical signals become less coherent.

Mistake 5: Treating AI Overview citations as guaranteed traffic

A citation may create visibility without producing a session. Assess citation presence as an awareness or authority indicator, then measure on-site results separately.

Mistake 6: Ignoring existing content

A refresh campaign may improve relevance, internal linking and conversion performance more efficiently than another batch of new articles. The best next page is sometimes the one already published.

An Operating Process for SEO Teams

If you manage a content programme, use a monthly or quarterly review cycle.

Monthly review

Focus on:

• Query and URL overlap.
• CTR changes.
• Brand versus non-brand performance.
• Pages with rising impressions and falling clicks.
• Organic conversion trends.
• New AI Overview observations.
• Broken or weak internal links.
• Freshness issues.

Quarterly review

Go deeper:

1. Rebuild the topic and intent map.
2. Review competing pages and SERP changes.
3. Score consolidation candidates.
4. Compare page-level and cluster-level performance.
5. Audit commercial paths from informational content.
6. Refresh content with declining relevance.
7. Review schema and technical indexation.
8. Reassign ownership for strategic topics.
9. Update the publishing calendar.
10. Record decisions and expected outcomes.

A repeatable process is more valuable than a one-off report. It gives you a consistent way to separate genuine ranking problems from changing search behaviour.

A Practical Template for an AI Overview Attribution Note

When reporting on a topic affected by AI Overviews, use language such as:

Organic clicks for this query group declined by 18% while impressions increased by 31%. Search Console does not currently provide enough evidence to identify whether AI Overview exposure caused the CTR change. Manual SERP checks show AI-generated summaries for several priority queries. At the same time, branded organic searches and assisted conversions increased. The current interpretation is that visibility has expanded while direct-click behaviour has become less predictable. We recommend improving the commercial journey, consolidating overlapping pages and monitoring the cohort for another reporting period.

This is more defensible than writing:

AI Overviews caused an 18% traffic decline.

The second statement may be true in part, but the available data does not prove it.

Key Takeaways for Marketers

The important points are straightforward, even though the reporting environment is not:

• Search Console measures search performance, not every form of search influence.
• Analytics measures tracked website interactions, not unclicked AI Overview exposure.
• A cited page may influence demand without receiving a visit.
• AI Overviews can reduce CTR while increasing brand awareness or query coverage.
• Keyword cannibalisation becomes harder to diagnose when several URLs contribute to one topic.
• Keyword overlap SEO should be assessed through intent and page purpose, not wording alone.
• Duplicate keyword targeting can weaken both rankings and attribution clarity.
• A content consolidation strategy should consider links, conversions, intent and evidence quality.
• Cluster-level reporting is often more reliable than URL-level reporting.
• Organic traffic should be reviewed alongside assisted conversions, branded demand and revenue.
• AI Overview observations are useful directional evidence, not perfect attribution.
• A disciplined publishing workflow helps prevent content overlap before it becomes a reporting issue.

Build a More Disciplined Publishing Operation With SEO Letters

AI Overviews have not made SEO measurement irrelevant. They have made simplistic measurement less reliable.

You still need to understand which topics matter, which pages should own them, where competitors have stronger coverage and whether published content contributes to qualified demand. You also need a practical way to research, write, link, publish and refresh content without creating a growing pile of overlapping URLs.

SEO Letters is built for that workflow. It combines keyword research, difficulty analysis, topical authority planning, site-gap analysis, structured AI article writing, internal linking, schema, image support, multilingual generation and direct publishing in one operating system. Its autonomous campaign scheduler can research, write and publish on a defined cadence, while content-refresh campaigns help maintain the pages you already rely on.

If you’re dealing with unstable rankings, unclear attribution or a growing keyword cannibalisation problem, start by mapping your search intents and assigning a clear URL to each one. Then use the SEO Letters app to turn that strategy into a repeatable publishing and refresh process.

For questions about your content workflow, campaign structure or site-wide overlap, use the rightbar as the contact path. The objective is not to produce more pages for the sake of volume. It is to create a coherent publishing operation that earns visibility, supports real journeys and gives your reporting team evidence they can actually defend.