We analysed 14,728 search queries across four intent categories and two devices. The result? 67.7% of organic search opportunity is being absorbed by AI Overviews and SERP features before it ever reaches a website. And your keyword research tool won't tell you.
Your keyword research tool says "how to poach eggs" has 5,400 monthly searches. It gives you a difficulty score, a traffic potential estimate, and a comforting green arrow suggesting this is a keyword worth targeting.
What it doesn't tell you is that the organic opportunity score for this query is 96.3 out of 100. Meaning: of those 5,400 people searching, the overwhelming majority will never reach a website. They'll get their answer from an AI Overview, a featured snippet, a People Also Ask box, or an inline video — all before they scroll far enough to see an organic result.
This isn't an edge case. Across 14,728 queries analysed in the Wayfinder AI Search Absorption Index, the average composite absorption score is 67.7. On a scale where 0 means an untouched organic SERP and 100 means organic results are effectively invisible, 67.7 is firmly in "High Absorption" territory. And in some categories, it's "Critical."
Here's what this post argues: Keyword research is broken because it measures demand without measuring accessibility. Search volume tells you how many people want an answer. It doesn't tell you whether they'll ever reach your site to find it. The gap between those two numbers — demand and accessible demand — is what we're calling absorption. And it's the single most important metric missing from your keyword strategy.
The standard SEO workflow hasn't changed in a decade. Find a keyword. Check the volume. Check the difficulty. Optimise a page. Publish. Wait. Pray.
The entire workflow rests on an unstated assumption: that search volume translates into click volume. That if 5,400 people search for "how to poach eggs" and you rank in the top three, you'll get a predictable share of those clicks based on published CTR curves.
That assumption was always shaky. Now it's outright wrong.
The problem isn't that CTR curves have shifted. It's that the SERP itself has been re-architected. Where a query once returned ten blue links and maybe an ad, it now returns an AI Overview, a featured snippet, a People Also Ask box with expandable answers, inline videos, shopping carousels, knowledge panels, and local packs. By the time the user reaches the first organic result, they've already encountered multiple opportunities to get what they need without clicking.
And then there's the AI Overview. Google's generative answer, synthesised from multiple sources and displayed at the top of the page. It doesn't just push organic results down. In many cases, it removes the need to click at all.
Here's what this looks like in practice. "How to poach eggs": 5,400 monthly searches, 96.3 absorption. "Best fuel injector cleaner": 22,200 searches, 95.9 absorption. "What is hackrf one": 10 searches, 96.7 absorption. The volume varies by three orders of magnitude. The absorption is near-total in every case.
Search volume and absorption are decoupled: high-volume navigational queries can have near-zero absorption while low-volume informational queries can be fully absorbed.
The mental model shift is this: stop asking "how many people search for this?" and start asking "how many people will actually reach a site for this?" The first number is what your keyword tool shows. The second is what determines your traffic. And the gap between them is getting wider every month.
The Absorption Index is a 0–100 score where higher means more organic opportunity lost. It combines three sub-indices:
The composite across all 14,728 queries: 67.7.
Headline absorption metrics across the 14,728-query dataset.
This is a directional metric, not a precise traffic forecast. We don't claim that a 67.7 score means exactly 67.7% of clicks are lost. What it means is that, on average, organic results face a severely compromised environment. The score tells you whether a keyword is worth targeting in the current SERP landscape — something no existing keyword tool does.
To put the click absorption number in context: FirstPageSage's 2026 benchmark puts position 1 organic CTR at 39.8% on a clean SERP1. On a SERP with an AI Overview, Seer Interactive found organic CTR fell to 0.61% — a 61% drop from 1.76% in June 20242. When we say click absorption is 70.0, we're saying the average query in our dataset faces a CTR environment dramatically worse than the clean-SERP baseline. The direction is unambiguous.
The 67.7 composite isn't just one problem. It's two problems layered on top of each other.
Problem one — the baseline: Before AI Overviews existed, SERP features already absorbed 41.6% of organic opportunity. Featured snippets, People Also Ask, inline videos, shopping results, knowledge panels, local packs — these features have been crowding the SERP for years. SEOs have been complaining about "SERP crowding" since PAA boxes exploded in 2019. The data confirms they were right.
Problem two — the compounding: Then AI Overviews arrived, and the composite jumped to 67.7. That's a 26.1-point delta from non-AI features alone.
The double hit: AI-only absorption plus non-AI SERP feature absorption.
It's not one punch. It's a one-two combo. First the SERP gets crowded with features that push your result down. Then the AI Overview arrives and answers the question before the user even needs to scroll. The non-AI features are the supporting actors that make the AI Overview the star.
The timeline matters here. Google introduced Search Generative Experience (SGE) as a Labs experiment in May 20233. By May 2024 it had been rebranded as AI Overviews and launched to all US users. By October 2024 it had expanded to 100+ countries. By early 2026, BrightEdge reported AI Overviews appearing on 48%+ of tracked queries, a 58% year-over-year increase4. The shift from experimental opt-in to default feature for billions of users took roughly twelve months.
And it's not slowing down. Google AI Mode, launched in March 2025 as a conversational search experience, represents the next phase: not just a summary at the top of the page, but a full conversational interface that keeps the user inside Google's ecosystem5.
The structural trend is clear. The SERP has been progressively crowded for a decade. AI Overviews have accelerated that crowding by an order of magnitude. And the next iteration — AI Mode — will likely accelerate it further.
If the headline number is 67.7, the category-level numbers tell a more alarming story. Because absorption isn't evenly distributed. It clusters heavily around specific intent types.
| Category | Composite | Visibility | Click | Intent | AI-Only |
|---|---|---|---|---|---|
| Informational | 76.4 | 92.3 | 84.3 | 63.3 | 52.2 |
| Commercial | 72.1 | 84.0 | 74.4 | 58.7 | 46.6 |
| Transactional | 52.0 | 59.5 | 46.9 | 53.4 | 24.2 |
| Navigational | 48.5 | 56.3 | 42.3 | 51.2 | 16.0 |
Absorption scores by intent category across all sub-indices.
Informational queries are in critical territory. A composite of 76.4 means the average informational query faces severe absorption. "How to poach eggs" (96.3), "how to hard boil eggs" (95.5), "what is autolyse in baking" — these are the queries where AI Overviews excel. They give step-by-step instructions, definitions, and explanations directly in the SERP. The user gets what they need without a click.
Commercial queries are dangerous. "Best fuel injector cleaner" (95.9), "best drawing tablet" (95.9), "lifelock vs aura" (95.9) — product comparisons and "best" roundups are heavily summarised by AI. The user gets a recommendation, and while some will still click to compare, the window is narrowing as AI Overviews get better at product comparison.
Transactional queries are moderate. At 52.0 composite, these are more defensible because the user has intent to buy. The AI can recommend a product, but the user still needs to check out, see pricing, verify availability. There's still a journey to complete.
Navigational queries are the defensive moat. At 48.5 composite, these resist absorption because the user explicitly wants a specific site. "Indeed login" (3.7), "capital one login" (3.7), "ziprecruiter login" (3.7) — the lowest absorption scores in the entire dataset. Users searching for a brand's login page aren't looking for an AI summary. They're looking for the login page.
The most and least absorbed queries in the dataset.
The strategic implication is immediate: your content mix matters. A site heavy on informational "how to" content is at far more risk than a site heavy on branded navigational queries. The category you optimise for determines how much of your keyword volume is actually accessible.
And it gets worse when you look at the funnel.
Smaller screens mean less real estate. AI features consume more of the viewport. The fold is higher. Organic results get pushed down faster.
Mobile composite: 69.1. Desktop composite: 66.4. The gap is consistent across every category.
| Category | Desktop | Mobile | Gap |
|---|---|---|---|
| Informational | 75.6 | 77.3 | +1.7 |
| Commercial | 70.8 | 73.4 | +2.6 |
| Transactional | 50.2 | 53.8 | +3.6 |
| Navigational | 46.0 | 51.0 | +5.0 |
Mobile absorption is consistently higher than desktop across every intent category.
The navigational gap is the most striking. On desktop, navigational queries average 46.0 composite — already the most defensible category. On mobile, that jumps to 51.0. Even the safest category gets measurably less safe on a smaller screen.
If your traffic skews mobile — and for most sites, mobile represents 50-60% of search volume6 — your aggregate risk is higher than the headline 67.7 suggests. The "mobile-first" SEO mantra, which Google forced on the industry with mobile-first indexing, now has a dark twin: mobile-first absorption.
The old keyword research formula is:
Volume × (1 ÷ Difficulty) = Opportunity
The new formula needs a third term:
Volume × (1 ÷ Difficulty) × (1 − Absorption) = Real Opportunity
Ahrefs gives you traffic potential, but it's based on aggregate position CTR curves, not per-query SERP feature analysis. It assumes a relatively clean SERP. It doesn't know that "how to poach eggs" has an AI Overview, a featured snippet, three PAA boxes, and two inline videos — all before the first organic result.
The old keyword formula ignores absorption; the new one includes it.
But there's a deeper pattern that keyword tools miss entirely, and it's about where in the purchase funnel your keywords sit.
When we mapped our dataset against purchase stage — awareness, consideration, decision, retention — a stark pattern emerged. Awareness-stage queries average 76.4 composite. Consideration-stage: 72.1. Decision-stage: 52.5. Retention-stage: 48.77.
The funnel is being eaten from the top down. Your top-of-funnel content — the "what is," "how to," and "best" queries that content teams love to target — is the most absorbed. Your bottom-of-funnel content — the branded, transactional, and retention queries — is the most defensible.
And here's the value paradox: medium-value keywords (the awareness and consideration content that "builds authority") average 76.4 composite. High-value keywords (the decision and conversion queries) average 64.3. The keywords businesses most want to rank for are actually more defensible than the keywords they invest heavily in content for. Medium-value queries skew awareness-stage; high-value queries skew decision-stage — which is why the paradox exists.
Absorption increases up-funnel: awareness-stage queries are most absorbed.
What teams should do:
Score your existing keyword portfolio for absorption. Add a third column to your keyword spreadsheet. If a query scores above 75, it's in critical territory. Above 50, it's high risk. Below 50, it's defensible.
Shift investment toward lower-absorption categories. Navigational, transactional, and branded queries resist absorption because the user wants a specific destination. Decision-stage and retention-stage content is more defensible than awareness-stage.
Deprioritise high-absorption informational terms where the AI is already winning. Not because they're worthless — they may still serve brand awareness and citation in AI Overviews — but because their direct traffic potential is minimal.
This post has mapped the double hit: SERP crowding plus AI Overviews, compounded by category divides and mobile effects. But the headline numbers only tell part of the story.
In Part 2, we dig into intent absorption — the sub-index where AI doesn't just push you down the page, it removes the need to click entirely. We'll look at which queries get fully absorbed by AI answers, which survive, and what that means for content strategy when the AI is literally designed to answer the queries your content targets.
In Part 3, we break down the sector-level analysis. Which industries face the most severe absorption? Which semantic content types are uninhabitable? And what does the purchase-stage data mean for how you allocate your content budget across the funnel?
The interactive Absorption Index Browser is live at /resources/tools/absorption-index-browser — explore the full dataset, filter by category, device, sector, and semantic cluster, and see how your keywords stack up.
If you want us to run the Absorption Index on your vertical, email consulting@wayfinderai.tools.
Keyword research is broken because it ignores absorption. It measures demand without measuring accessibility. And in the AI search era, that gap is the difference between a content strategy that drives traffic and one that drives impressions inside Google's ecosystem.
The SERP has changed in two phases. First, a decade of feature creep: snippets, PAA, inline media, shopping carousels. Then, the AI acceleration: AI Overviews, AI Mode, and whatever comes next. The baseline absorption from non-AI features is 41.6%. The total is 67.7%. And in informational queries — the content type most SEO strategies are built around — it's 76.4.
The data is the map. The action is up to you.
FirstPageSage, "Google Click-Through Rates (CTRs) by Ranking Position in 2026," May 2025. ↩
Seer Interactive, "AIO Impact on Google CTR: September 2025 Update," November 2025. Study of 3,119 informational queries across 42 organisations. ↩
Ahrefs, "Google AI Overviews: All You Need to Know," May 2026. Comprehensive timeline of SGE → AI Overviews → AI Mode. ↩
BrightEdge, "AI Overviews at the One-Year Mark," February 2026. ↩
Search Atlas, "Google AI Mode: How It Works, Features, SEO Impact & Access Guide," May 2026. ↩
Industry estimates: mobile search volume globally ~60%. Statista, Similarweb. ↩
Wayfinder AI Search Absorption Index, June 2026. 14,728 queries merged with keyword catalog enrichment (category, semantic_cluster, purchase_stage, value). ↩
Next in series
Intent absorption: the threat keyword tools cannot see.
