Google’s AI Overviews Just Turned “How-To” Into a Zero-Click Commodity—The 7-Step Rewrite That Still Gets You Clicked (and the one thing most guides forget to publish)
AI Overviews don’t just compress your steps—they compress your business model. Here’s what the best data says about the click collapse, and how to rewrite guides so the summary isn’t the finish line.
By TheMurrow Editorial
March 24, 2026
Key Points
- 1Track the shift: AI Overviews cut traditional-result clicks nearly in half, and source-link clicks inside the box hover around ~1%.
- 2Reframe the goal: rankings matter less when the SERP satisfies intent—being cited inside AI Overviews can still lift CTR meaningfully.
- 3Rewrite for indispensability: add templates, visuals, edge cases, and decision guidance so the overview can’t “finish” the job without a click.
A decade ago, the web’s great bargain was simple: publishers wrote the instructions; Google sent the readers. Type a question into a search box—how to change a tire, how to write a cover letter, how to clean a cast-iron pan—and you were ushered to someone else’s site where the real work happened.
That bargain is fraying, and not because people stopped asking practical questions. The shift is happening because Google increasingly answers them before anyone clicks. The most visible expression of that shift is AI Overviews—Google’s AI-generated summaries that often appear above traditional results when Google decides the feature is “additive to classic Search,” according to its own Search Central documentation.
For “how-to” publishers, the danger isn’t hypothetical. It’s measurable, and it’s structural. The guide genre is built to be condensed: a finite set of steps, a predictable set of tools, a clean ending. When a search engine can satisfy that intent on the results page, the rest of the web starts to look like optional reading.
“The how-to guide is the internet’s most useful genre—and the easiest one to compress.”
— — TheMurrow Editorial
AI Overviews didn’t invent zero-click. They scaled it.
Google has been moving toward “satisfy on SERP” behavior for years: featured snippets, knowledge panels, weather boxes, sports scores, flights, product carousels. The common thread is that the searcher’s need gets met without leaving Google.
The SparkToro + Datos “Zero-Click Search Study” (2024) put a blunt number on the trend. For every 1,000 U.S. Google searches, only 360 clicks go to the “open web,” meaning external sites rather than Google-owned destinations. That is not an AI-specific phenomenon. It’s a platform incentive: keep the user, shorten the journey, reduce friction.
AI Overviews amplify that incentive because they can handle messy queries that used to require reading. A featured snippet might answer a single line. AIO can offer a summary, caveats, and a list of steps—often enough for a user to stop.
Google’s public framing is careful. In Search Central documentation, the company describes AI Overviews as a “jumping off point” that helps people explore, with supporting links to sources. The language matters: it implies AIO supplements the web rather than replaces it.
Publishers see the other side. If the “jumping off point” becomes the finish line, the click is no longer the default outcome—it’s a bonus.
The SparkToro + Datos “Zero-Click Search Study” (2024) put a blunt number on the trend. For every 1,000 U.S. Google searches, only 360 clicks go to the “open web,” meaning external sites rather than Google-owned destinations. That is not an AI-specific phenomenon. It’s a platform incentive: keep the user, shorten the journey, reduce friction.
AI Overviews amplify that incentive because they can handle messy queries that used to require reading. A featured snippet might answer a single line. AIO can offer a summary, caveats, and a list of steps—often enough for a user to stop.
Google’s public framing is careful. In Search Central documentation, the company describes AI Overviews as a “jumping off point” that helps people explore, with supporting links to sources. The language matters: it implies AIO supplements the web rather than replaces it.
Publishers see the other side. If the “jumping off point” becomes the finish line, the click is no longer the default outcome—it’s a bonus.
360
Out of every 1,000 U.S. Google searches, only 360 clicks go to the open web (SparkToro + Datos, 2024).
Why “how-to” gets hit first
“How-to” content is uniquely vulnerable because it tends to be:
- Procedural (steps can be summarized)
- Informational (queries often don’t require a transaction)
- Finite (there’s a sense of completion)
Those traits make it ideal for AI summarization and for long-standing zero-click behavior. The uncomfortable truth: many how-to queries were already low-click even before generative summaries. AI Overviews don’t create the problem so much as deepen it.
- Procedural (steps can be summarized)
- Informational (queries often don’t require a transaction)
- Finite (there’s a sense of completion)
Those traits make it ideal for AI summarization and for long-standing zero-click behavior. The uncomfortable truth: many how-to queries were already low-click even before generative summaries. AI Overviews don’t create the problem so much as deepen it.
Why how-to guides are easy to compress
- ✓Procedural (steps can be summarized)
- ✓Informational (queries often don’t require a transaction)
- ✓Finite (there’s a sense of completion)
The click collapse: what the best data says
The strongest evidence so far comes not from SEO tool vendors selling dashboards, but from behavioral measurement.
The Pew Research Center analyzed browsing behavior of 900 U.S. adults in its KnowledgePanel Digital sample, tracking activity from March 1–31, 2025. Pew’s work is valuable because it observes what people did, not what they said they do.
In widely reported coverage of Pew’s findings:
- Search pages without AI Overviews saw a 15% click rate to traditional results.
- Search pages with AI Overviews saw an 8% click rate—nearly half.
- Only about 1% of AI Overview views produced a click on a source link inside the AI box.
Ars Technica, summarizing Pew’s results, highlighted the core implication: AI answers can be the end of the journey, not the start.
Seer Interactive’s tracking offers a different angle: what happens to click-through rates over time as AIO presence grows. Seer analyzed 3,119 informational queries from June 2024 to September 2025 across 42 organizations, reporting that:
- Organic CTR for queries with AIO fell from 1.76% (June 2024) to 0.61% (Sept 2025), a 61% decline.
- Paid CTR on AIO queries fell 68% over the same period.
Those numbers are not the death of search traffic. They are, however, a re-pricing of attention—especially for publishers whose business model assumes that being the best answer earns the click.
The Pew Research Center analyzed browsing behavior of 900 U.S. adults in its KnowledgePanel Digital sample, tracking activity from March 1–31, 2025. Pew’s work is valuable because it observes what people did, not what they said they do.
In widely reported coverage of Pew’s findings:
- Search pages without AI Overviews saw a 15% click rate to traditional results.
- Search pages with AI Overviews saw an 8% click rate—nearly half.
- Only about 1% of AI Overview views produced a click on a source link inside the AI box.
Ars Technica, summarizing Pew’s results, highlighted the core implication: AI answers can be the end of the journey, not the start.
Seer Interactive’s tracking offers a different angle: what happens to click-through rates over time as AIO presence grows. Seer analyzed 3,119 informational queries from June 2024 to September 2025 across 42 organizations, reporting that:
- Organic CTR for queries with AIO fell from 1.76% (June 2024) to 0.61% (Sept 2025), a 61% decline.
- Paid CTR on AIO queries fell 68% over the same period.
Those numbers are not the death of search traffic. They are, however, a re-pricing of attention—especially for publishers whose business model assumes that being the best answer earns the click.
15% → 8%
Pew reported traditional-result click rates nearly halved when AI Overviews appeared (March 2025 behavioral tracking).
~1%
Only about 1% of AI Overview views produced a click on a source link inside the AI box (Pew, widely reported).
1.76% → 0.61%
Seer found organic CTR on AIO queries fell 61% from June 2024 to Sept 2025 across 3,119 informational queries.
“AI Overviews don’t just change rankings. They change whether a ranking matters.”
— — TheMurrow Editorial
A crucial nuance: citation can still help
Seer’s dataset also contained a lifeline: sites cited inside AI Overviews saw 35% higher organic CTR and 91% higher paid CTR relative to non-cited sites in their analysis.
That doesn’t negate the overall decline. It does clarify the new competition. The old goal was position one. The emerging goal, for many informational queries, is: become the source the machine is willing to quote.
That doesn’t negate the overall decline. It does clarify the new competition. The old goal was position one. The emerging goal, for many informational queries, is: become the source the machine is willing to quote.
Key Insight
The old SEO target was #1. The new target (for many informational queries) is: be the source the machine chooses to cite.
Google’s official stance: no secret door, no special tag
Publishers love a workaround. When a platform changes, the instinct is to hunt for a new schema type, a hidden setting, a “magic switch” that restores the old equilibrium.
Google’s documentation offers little comfort on that front. Search Central states that there are no additional requirements to appear in AI Overviews or AI Mode beyond being indexed and eligible for a snippet, alongside meeting technical requirements and policies.
Read that again. Google is saying: there is no special opt-in checklist for AIO visibility. Your content competes under the same general rules as classic search—crawlability, quality, relevance, and the many signals Google already uses.
From a publisher perspective, that has two implications.
First, technical SEO remains necessary but insufficient. Being fast, clean, and structured won’t automatically earn you a citation.
Second, the advantage shifts to editorial choices: what you publish, how you frame it, what you can prove, and whether your page provides something the overview cannot finish.
Google’s documentation offers little comfort on that front. Search Central states that there are no additional requirements to appear in AI Overviews or AI Mode beyond being indexed and eligible for a snippet, alongside meeting technical requirements and policies.
Read that again. Google is saying: there is no special opt-in checklist for AIO visibility. Your content competes under the same general rules as classic search—crawlability, quality, relevance, and the many signals Google already uses.
From a publisher perspective, that has two implications.
First, technical SEO remains necessary but insufficient. Being fast, clean, and structured won’t automatically earn you a citation.
Second, the advantage shifts to editorial choices: what you publish, how you frame it, what you can prove, and whether your page provides something the overview cannot finish.
The “jumping off point” promise vs. user behavior
Google says AIO encourages exploration through supporting links. Pew’s panel evidence suggests exploration often doesn’t happen. Both can be true. AI Overviews may help some users discover sources, while also giving many users enough to stop.
Publishers don’t need to accuse Google of bad faith to recognize the outcome: a feature designed to reduce effort will, predictably, reduce clicks.
Publishers don’t need to accuse Google of bad faith to recognize the outcome: a feature designed to reduce effort will, predictably, reduce clicks.
How-to content is being commoditized—because it was built to be
“How-to” became the workhorse of SEO for a reason. It matches user intent cleanly. It attracts links. It is evergreen. It can be updated. It can be templated without obvious harm.
That last point is now the genre’s weakness.
A well-optimized guide often looks like every other well-optimized guide:
- a short intro
- a tools list
- steps in order
- a brief FAQ
From a reader standpoint, that uniformity is helpful. From an AI standpoint, it’s a dream. The content is predictable, extractable, and easy to summarize.
Seer’s reporting adds an important framing: AI Overviews tend to appear on query types already predisposed to lower clicks—often informational, quick-answer-like searches. In other words, AIO isn’t necessarily targeting your best-performing pages at random. It’s concentrating on the kinds of queries where Google can complete the job without the web.
For editorial teams, the hard question isn’t “How do we outrank everyone?” It’s “How do we publish work that can’t be finished on the results page?”
That last point is now the genre’s weakness.
A well-optimized guide often looks like every other well-optimized guide:
- a short intro
- a tools list
- steps in order
- a brief FAQ
From a reader standpoint, that uniformity is helpful. From an AI standpoint, it’s a dream. The content is predictable, extractable, and easy to summarize.
Seer’s reporting adds an important framing: AI Overviews tend to appear on query types already predisposed to lower clicks—often informational, quick-answer-like searches. In other words, AIO isn’t necessarily targeting your best-performing pages at random. It’s concentrating on the kinds of queries where Google can complete the job without the web.
For editorial teams, the hard question isn’t “How do we outrank everyone?” It’s “How do we publish work that can’t be finished on the results page?”
“If your guide can be reduced to seven bullet points, Google will eventually reduce it to seven bullet points.”
— — TheMurrow Editorial
The real business risk: the mid-funnel collapse
“How-to” traffic often sits in the middle of a publisher’s funnel. It’s how people discover a brand before they subscribe, buy, or return. When those entry points shrink, the long-term damage shows up months later: fewer returning visitors, fewer email signups, fewer conversions, less ad inventory.
AI Overviews don’t just siphon a pageview. They can remove the first handshake.
AI Overviews don’t just siphon a pageview. They can remove the first handshake.
What actually breaks when clicks disappear
Mid-funnel discovery shrinks first, then returning visitors, email signups, conversions, and ad inventory follow months later.
The new playbook: avoid the trigger, win the citation, or make the click essential
Publishers are not powerless here, but the tactics are more editorial than mechanical. Seer’s nuance—citations correlating with higher CTR—points toward a strategy that treats AIO like a new layer of distribution rather than a temporary bug.
1) Avoid AIO triggers (selectively)
Not every query is equally likely to produce an AI Overview. Seer notes AIO clusters around informational queries with quick-answer behavior. That suggests one path: aim content toward needs that are harder to satisfy in a summary.
Examples of “hard to summarize” angles include:
- Decision-making under constraints (budgets, edge cases, trade-offs)
- Comparisons that require context and preferences
- Original reporting: interviews, data, first-hand testing
- Local or time-sensitive nuance where answers vary
A generic “how to write a résumé” guide can be condensed. A piece that compares résumé formats by industry, seniority, and applicant-tracking systems—with examples and failure modes—is harder to finish in one box.
Examples of “hard to summarize” angles include:
- Decision-making under constraints (budgets, edge cases, trade-offs)
- Comparisons that require context and preferences
- Original reporting: interviews, data, first-hand testing
- Local or time-sensitive nuance where answers vary
A generic “how to write a résumé” guide can be condensed. A piece that compares résumé formats by industry, seniority, and applicant-tracking systems—with examples and failure modes—is harder to finish in one box.
Angles that are harder for AIO to finish
- ✓Decision-making under constraints (budgets, edge cases, trade-offs)
- ✓Comparisons that require context and preferences
- ✓Original reporting: interviews, data, first-hand testing
- ✓Local or time-sensitive nuance where answers vary
2) Win citations: write like a source, not a content farm
Google frames AIO as a summary with supporting links. Pew’s finding that only ~1% click those links is sobering, but Seer’s citation lift suggests the remaining clicks are valuable—and that being cited might be one of the few scalable ways to stay visible.
To increase the odds of being cited, favor:
- Clear, verifiable claims (definitions, numbers, explicit steps)
- Tight sectioning with descriptive subheads (easy for systems to parse)
- Distinctive expertise: demonstrated experience rather than generic advice
- Specificity that other pages don’t offer (edge cases, warnings, constraints)
Google’s documentation doesn’t promise how citation selection works. Still, the logic of machine summarization is straightforward: systems prefer text that is explicit, attributable, and easy to extract without distortion.
To increase the odds of being cited, favor:
- Clear, verifiable claims (definitions, numbers, explicit steps)
- Tight sectioning with descriptive subheads (easy for systems to parse)
- Distinctive expertise: demonstrated experience rather than generic advice
- Specificity that other pages don’t offer (edge cases, warnings, constraints)
Google’s documentation doesn’t promise how citation selection works. Still, the logic of machine summarization is straightforward: systems prefer text that is explicit, attributable, and easy to extract without distortion.
Make your guide cite-worthy
- ✓Clear, verifiable claims (definitions, numbers, explicit steps)
- ✓Tight sectioning with descriptive subheads
- ✓Distinctive expertise (demonstrated experience)
- ✓Specificity (edge cases, warnings, constraints)
3) Create a reason to click that AIO can’t provide
If the overview gives the steps, the click needs to offer something else: depth, tools, proof, or reassurance.
Practical ways to do that:
- Offer downloadable templates or checklists that require a visit
- Include interactive elements (calculators, quizzes, decision trees)
- Provide visual walkthroughs or annotated screenshots
- Add case studies and real outcomes rather than generic instruction
AI can summarize the “what.” Readers still click for the “show me” and “help me do it.”
Practical ways to do that:
- Offer downloadable templates or checklists that require a visit
- Include interactive elements (calculators, quizzes, decision trees)
- Provide visual walkthroughs or annotated screenshots
- Add case studies and real outcomes rather than generic instruction
AI can summarize the “what.” Readers still click for the “show me” and “help me do it.”
Editor’s Note
If the SERP can deliver the steps, your page has to deliver proof, tools, visuals, and judgment—things a summary can’t complete.
Case studies you can recognize: what survives when summaries take the basics
We don’t need confidential analytics to see the pattern. The web has long rewarded the sites that add something beyond instructions.
The “commodity guide” vs. the “lived-in guide”
A commodity guide explains the steps to, say, home Wi‑Fi troubleshooting. A lived-in guide explains the steps and the messy realities:
- Which fixes are safe to try first
- When to stop and call your ISP
- How to tell whether the router is dying
- What changes if you live in an apartment building
- A checklist for narrowing the problem in ten minutes
An AI Overview can capture the basic sequence. It struggles with the practical triage that reflects experience.
- Which fixes are safe to try first
- When to stop and call your ISP
- How to tell whether the router is dying
- What changes if you live in an apartment building
- A checklist for narrowing the problem in ten minutes
An AI Overview can capture the basic sequence. It struggles with the practical triage that reflects experience.
Two formats, two fates
Before
- Commodity guide (steps only)
- generic sequence
- clean finish
After
- Lived-in guide (triage + reality)
- safe-first fixes
- stop points
- edge cases
- fast checklist
The template that holds attention: “best of,” “mistakes,” “what I wish I knew”
Publishers who rely on classic how-to formats can reframe without abandoning utility. Instead of “How to negotiate rent,” consider:
- “Rent negotiation scripts for three scenarios”
- “Seven mistakes that get you a fast no”
- “What to prepare before you ask—documents and timing”
Those approaches don’t dodge usefulness. They reintroduce judgment, voice, and specificity—elements that summaries often flatten.
- “Rent negotiation scripts for three scenarios”
- “Seven mistakes that get you a fast no”
- “What to prepare before you ask—documents and timing”
Those approaches don’t dodge usefulness. They reintroduce judgment, voice, and specificity—elements that summaries often flatten.
The citation-friendly format: crisp answers with depth behind them
If your goal includes being cited, structure helps. Put the precise answer upfront, then expand.
A strong pattern:
1. A one- or two-sentence answer (cite-worthy)
2. A short list of steps (extractable)
3. The nuance: edge cases, examples, warnings (click-worthy)
Google calls AIO a “jumping off point.” Give it a ramp that points back to you.
A strong pattern:
1. A one- or two-sentence answer (cite-worthy)
2. A short list of steps (extractable)
3. The nuance: edge cases, examples, warnings (click-worthy)
Google calls AIO a “jumping off point.” Give it a ramp that points back to you.
A citation-friendly guide structure
- 1.A one- or two-sentence answer (cite-worthy)
- 2.A short list of steps (extractable)
- 3.The nuance: edge cases, examples, warnings (click-worthy)
A fair counterpoint: users benefit—and publishers must adapt anyway
It’s tempting to frame AI Overviews as a simple theft machine. The better critique is more uncomfortable: many users genuinely want faster answers, and search engines compete on speed and convenience.
Pew’s behavioral finding that users may be more likely to end their browsing session after encountering an AI summary (reported in secondary coverage as 26% vs. 16%) can be read two ways. Publishers read it as lost traffic. Users might read it as success: the task is done.
Google’s own documentation emphasizes that AIO appears when its systems decide the feature is “additive.” That implies Google sees a meaningful user benefit—at least in aggregate.
The question for publishers, then, isn’t whether AI Overviews should exist. The question is what kind of publishing is still rewarded when basic synthesis becomes free.
The answer is not to abandon how-to. It’s to stop treating how-to as an SEO template and start treating it as editorial craft.
Pew’s behavioral finding that users may be more likely to end their browsing session after encountering an AI summary (reported in secondary coverage as 26% vs. 16%) can be read two ways. Publishers read it as lost traffic. Users might read it as success: the task is done.
Google’s own documentation emphasizes that AIO appears when its systems decide the feature is “additive.” That implies Google sees a meaningful user benefit—at least in aggregate.
The question for publishers, then, isn’t whether AI Overviews should exist. The question is what kind of publishing is still rewarded when basic synthesis becomes free.
The answer is not to abandon how-to. It’s to stop treating how-to as an SEO template and start treating it as editorial craft.
AI Overviews: the tension
Pros
- +Faster answers
- +less friction
- +tasks completed sooner
Cons
- -Fewer publisher clicks
- -weaker mid-funnel discovery
- -less incentive to fund expertise and reporting
Where this leaves the open web
The web doesn’t die because fewer people click on “how to boil eggs.” It erodes when the economic logic of publishing breaks: fewer readers arriving, fewer subscriptions, fewer ads, fewer budgets for reporting and expertise.
The most telling statistic in this entire debate may be Pew’s: only about 1% of AI Overview views generated a click on a source link inside the box. Supporting links exist, but they are not behaving like the old referral engine.
Still, Seer’s citation lift suggests a practical path forward: treat AI Overviews as a new distribution layer where being cited can meaningfully improve performance, even as overall CTR declines.
Publishers should be clear-eyed. Google will keep pushing toward faster answers. The “how-to” genre will keep being compressed. The web’s strongest defense isn’t outrage. It’s differentiation—work that earns trust, offers proof, and gives readers a reason to come back.
The next era of search traffic may belong less to the pages that explain and more to the pages that demonstrate.
The most telling statistic in this entire debate may be Pew’s: only about 1% of AI Overview views generated a click on a source link inside the box. Supporting links exist, but they are not behaving like the old referral engine.
Still, Seer’s citation lift suggests a practical path forward: treat AI Overviews as a new distribution layer where being cited can meaningfully improve performance, even as overall CTR declines.
Publishers should be clear-eyed. Google will keep pushing toward faster answers. The “how-to” genre will keep being compressed. The web’s strongest defense isn’t outrage. It’s differentiation—work that earns trust, offers proof, and gives readers a reason to come back.
The next era of search traffic may belong less to the pages that explain and more to the pages that demonstrate.
“The next era of search traffic may belong less to the pages that explain and more to the pages that demonstrate.”
— — TheMurrow Editorial
T
About the Author
TheMurrow Editorial is a writer for TheMurrow covering how-to / guides.
Frequently Asked Questions
What are Google AI Overviews, exactly?
AI Overviews (AIO) are AI-generated summaries that can appear above traditional results when Google’s systems determine the feature is “additive to classic Search,” according to Google Search Central documentation. Google frames them as a “jumping off point” that includes supporting links to sources, though users often don’t click those links.
Do AI Overviews really reduce clicks to websites?
Evidence suggests yes. Pew Research Center tracked browsing behavior of 900 U.S. adults in March 2025 and found traditional-result click rates dropped from 15% on pages without AIO to 8% on pages with AIO. Secondary reporting also noted that only about 1% of AIO views led to a click on a source link inside the AI box.
Why are “how-to” guides getting hit harder than other content?
“How-to” content is often procedural and finite—perfect for summarization. Many of these searches are informational, and informational queries already skew toward lower-click behavior. Seer Interactive’s analysis suggests AIO tends to show on query types predisposed to quick answers, meaning the format itself invites on-page satisfaction.
Is there a technical trick to get featured in AI Overviews?
Google says no special requirements exist beyond being indexed and eligible for a snippet, plus meeting general technical and policy requirements. That means there’s no official “AIO schema switch.” Competitive advantage is more likely to come from editorial execution—clarity, specificity, credibility, and usefulness that stands out.
If clicks are falling, is SEO still worth it for publishers?
SEO still matters, but expectations need updating. Seer Interactive reported organic CTR declines on AIO queries (from 1.76% to 0.61% between June 2024 and Sept 2025), yet also found that being cited in AIO correlated with higher CTR. Visibility may increasingly come from citations and brand recognition, not just rankings.
What should publishers do with existing how-to content?
Start by identifying which guides target quick-answer informational queries—those are most likely to be summarized. Then rewrite with added value AIO can’t complete: real examples, edge cases, interactive tools, checklists, and decision guidance. Structure content so it’s easy to cite (clear answers) but still worth clicking (depth, demonstration, proof).
