If your organic traffic is down, you're not imagining it.

As of March 2026, Google AI Overviews appear on 48% of all search queries, up from 34.5% in December. On informational queries, they appear 70% of the time. Position 1 CTR on queries with an AI Overview has dropped to 8-12%, compared to 28-34% on queries without one. The Pew Research Center's controlled study of 68,000 queries found a 46.7% relative decline in clicks when an AI Overview was present.

The search landscape changed. Not gradually, it changed fast, and most SEO teams are still running a playbook built for conditions that no longer exist.

Here's the important distinction: SEO isn't dead. The brands winning search in 2026 are growing their organic visibility, getting cited inside AI Overviews, and driving higher-intent traffic than they got from the old keyword-volume model. But they're doing something fundamentally different from what most teams are doing.

The shift is from keyword optimisation to topic authority. Those are different strategies producing different results. Understanding the distinction and what to do about it, is what this issue is for.

What the data actually says and what it doesn't

The headline numbers are alarming but incomplete. Yes, CTR has dropped significantly on AI Overview-affected queries. Yes, HubSpot estimates 70-80% traffic decline on some content categories. Yes, Chegg's revenue fell 24% year-on-year and the company sued Google explicitly citing AI Overviews. The damage to commodity informational content is real and ongoing.

But here's what the same data also shows: brands cited inside AI Overviews see 35% more organic clicks than non-cited brands on the same SERP. Branded queries with AI Overviews actually see an 18% CTR increase. Visitors who do click through from AI Overview-adjacent results convert at measurably higher rates than standard search visitors — because Google has already qualified their intent upstream.

The search funnel didn't disappear. It restructured. Google is now doing the top-of-funnel filtering — answering informational queries directly — and sending the clicks it does pass through to sources it considers authoritative. The question isn't whether you're getting fewer clicks. It's whether you're positioned to receive the clicks that remain.

The brands losing the most are the ones whose SEO strategy was built on volume: high-output content targeting informational keywords, optimised for keyword density, published at scale. That model is functionally over. Google's AI can synthesise the same information from multiple sources without directing a single click to any of them.

The brands gaining are the ones who built genuine topical authority — depth and consistency in a defined subject area — before the restructuring happened. They're being cited inside AIOs. Their brand appears in AI answers. Their traffic is lower in volume and higher in conversion quality.

→ Takeaway: Run your current content against this question: does our site demonstrate sustained, interconnected expertise in a defined topic area — or do we have a collection of individual pages targeting individual keywords with no structural relationship between them? The answer tells you which model you're running, and which results to expect.

What topical authority actually means and how to build it

Topical authority is not a new concept. What's changed is how consequential it's become and how precisely search systems can evaluate it.

In simple terms: a keyword is a moment. A topic is a territory. Keyword optimisation asks "can we rank for this query?" Topic authority asks "does our site own this subject area in the eyes of search systems?" The second question is harder to answer and harder to game — which is why the brands that are genuinely building it have an increasingly durable competitive position.

The practical architecture is the content cluster: a pillar page covering a topic comprehensively, surrounded by cluster pages covering every significant subtopic, linked internally in a coherent structure. The pillar signals to Google what territory you're claiming. The cluster pages signal the depth and consistency of your expertise within it. Together they produce authority that individual pages can't — and the compounding effect builds over 12+ months.

What content earns citations inside AI Overviews? The research is consistent: structured, authoritative content that directly answers the query in the first paragraph, uses clear headings, includes specific data points or original insight, and comes from a domain with established topical depth. Generic content optimised for keywords performs poorly. Extractable expertise performs well.

The distinction that matters: there is ranking content and there is citable content. Ranking content is optimised around keywords — and has become increasingly homogenised as everyone has followed the same formula. Citable content is original, specific, and useful enough that other sources — and AI systems — want to reference it. The goal in 2026 is to produce citable content. Ranking follows from it, not the other way around.

The practical implication for most teams: stop measuring success by keyword rankings alone. The new KPIs include share of voice in AI citations, branded search volume trend, and assisted conversion value from organic. A team optimising for citations will produce different content — with different structure, different specificity, and different editorial standards — than a team optimising for keyword rankings.

→ Takeaway: Pick one topic area your brand should own. Map every subtopic within it. Audit how many of those subtopics have a page on your site — and whether those pages are linked to each other. The gaps in that map are your content priority list. Fill them in order of business relevance, not search volume.

The practical shift from old SEO to new SEO, without abandoning what still works

The temptation when a channel changes this significantly is to overreact in one direction or another — either to declare SEO dead and stop investing, or to chase every new optimisation framework without a coherent strategy. Neither helps.

Here's what the data and the winning case studies actually support:

Protect transactional and commercial intent pages first. AI Overviews appear on only 4% of e-commerce queries and 12% of transactional queries — far less than the 70%+ on informational content. Your product pages, category pages, and commercial landing pages are significantly less disrupted than your blog content. Invest in their technical SEO, structured data, and conversion quality before anything else.

Audit your informational content ruthlessly. The "what is X" pages, the basic explainers, the definition articles — these are the most exposed to AI Overview cannibalisation. Either upgrade them to a standard that earns AIO citation (original data, specific expertise, clear structure), or consolidate them into fewer, deeper pieces. Thin informational content that no longer earns clicks and doesn't earn citations is negative equity.

Build for the platforms getting cited. Reddit holds 21% of AIO citations. YouTube is the most-cited source across LLM answers overall. LinkedIn is the most-cited domain for professional and B2B queries across AI systems. If your brand is absent from the platforms Google is actively pulling citations from, you're structurally disadvantaged regardless of how good your owned content is. This isn't about social media strategy — it's about citation footprint.

Invest in what AI can't synthesise. Original research. Proprietary data. First-hand experience. Exclusive insight. These are the content types that AI Overviews cannot adequately replicate from aggregated sources — because they don't exist anywhere else. A study you conducted, a dataset you built, a perspective that comes from doing the thing rather than writing about it. This is the structural protection the winning publishers have found, and it's available to every brand.

→ Takeaway: Identify five questions your target customer asks that your brand should definitively answer. Check whether you have a page for each one that leads with a clear, direct answer in the first paragraph, uses structured headings, and includes a specific data point or original claim that no competitor has. If not, those five pages are your Q3 content priority.

Why less traffic might mean more revenue and how to measure it honestly

Here's the counterintuitive finding buried inside the AI Overviews data that most teams haven't processed yet.

Visitors who click through from AI Overview-adjacent results convert at significantly higher rates than standard organic visitors. The mechanism is straightforward: Google has already answered the basic question. The people who still click through are the ones who want more depth, more specificity, or direct engagement — not the ones who were satisfied by a summary. That's a fundamentally more qualified visitor.

This creates a measurement problem for teams still using traffic volume as their primary organic performance metric. If traffic falls 30% but conversion rate doubles, you have more revenue from fewer sessions. Your analytics report looks worse. Your business is performing better.

The shift this requires is from traffic-first measurement to intent-quality measurement. The metrics that matter in 2026: branded search volume trend (are people specifically looking for you?), organic assisted conversion value (what is organic's role in the full conversion path?), and citation share (are you appearing in the AI answers your customers are receiving?).

Teams that restructure their organic KPIs around these signals will make fundamentally different content decisions than teams still chasing keyword volume. They'll publish less, invest more per piece, and prioritise depth over breadth. The output looks different. The business results are better.

→ Takeaway: Pull your organic traffic in GA4 and look at conversion rate by landing page alongside traffic volume. Find your highest-converting organic pages. Those are your best evidence of what high-intent organic visitors look like — and what content type attracts them. Replicate that structure and depth across your topic cluster.

Ahrefs' AI Overview Tracker (or Semrush's AIO report if you're already on Semrush).

The specific use: enter your 20 most important target keywords and check which ones are triggering AI Overviews. Then check whether your domain is being cited inside those AIOs. The gap between "we rank for this" and "we're cited in the AIO for this" is your single most actionable optimisation target in 2026.

For brands not on paid SEO tools: Google Search Console now surfaces AI Overview impression data for some accounts. Check your Performance report for the "Search type: AI Overviews" filter — if it's available for your account, this is free baseline data for where your content is and isn't being surfaced.

Pick one topic your brand should own. Map the cluster.

Write down the central topic. Then write down every significant subtopic beneath it — every question a customer might have, every adjacent concept, every related decision they'd need to make.

Count how many of those subtopics have a page on your site. Count how many of those pages are linked to each other.

That map tells you exactly where your topical authority gaps are. Fill them in order of business relevance. Don't start with the highest search volume — start with the question your best customer asks most often that your site doesn't currently answer well.

That's the Q3 content brief. It takes 30 minutes to build and produces a clearer content strategy than most SEO audits.