Google has confirmed that for some queries AI Overviews may dynamically expand to its full state, replacing the AI Overview snippet with a full detailed response. This not only removes the “Show more” button in the AI Overview but it also pushes down the core search results that searchers were able to see without having to always scroll down the page.

Now the Google Search results look more like a full AI answer, more like how AI Mode works, versus how Google has typically shown AI responses within Search.

What it looks like. For some queries, Google will fully expand the AI Overview response, so the need to click on the “Show more” button will not be needed. Here is a GIF of it in action:

If users have already begun scrolling to view content below the AI Overview, Google cancels the expansion so that the user doesn’t lose their reading position.

What Google said. A Google spokesperson told us:

• “For some queries, AI Overviews may dynamically expand for topics where our systems determine it’s most useful for people.”

But as you can imagine, Google said this is a good thing. Google added:

• “Our research has shown that with this dynamic experience, users find Search more helpful and engage deeper in follow-up exploration.”

More details. Chris Long spotted this earlier and posted on X, he wrote, “Holy moly SEOs: Google is now revealing the AI Mode prompt box BY DEFAULT for some queries.” “This is how Google continues to push AI Mode as the default search experience,” he added.

Here is his post:

Why we care. If you thought AI Overviews take away clicks, just wait for this to roll out to your queries. This should lead to even fewer and fewer clicks from Google Search. Google earlier this week added link carousels for developing topics to AI Mode, like it did for AI Overviews. But I doubt that will help all that much for most queries that show this fully expanded answer.

Google did not comment on what percentage of queries this impacts and if it will continue to expand to more queries over time. But you’d have to imagine it will. We all expect AI Mode to take over Google Search at some point in time, and Google has been taking baby steps in that direction.

]]> google-ai-overview-no-show-more-Clm4zUw7 https://searchengineland.com/greg-finn-on-20-years-of-ppc-from-exact-match-to-ai-mode-486188 Fri, 28 Aug 2026 16:46:50 +0000 https://searchengineland.com/?p=486188 Greg Finn reflects on 20 years of PPC, from exact match and Smart Bidding to automation, AI Mode and why authenticity will matter more than ever.]]>

Cypress North co-founder Greg Finn has spent more than two decades watching search marketing evolve from a hands-on, highly granular discipline into an industry increasingly shaped by automation, AI and algorithms.

In a conversation celebrating Search Engine Land’s 20th anniversary, Finn reflected on his early days in PPC and SEO, the rise of Google Ads automation, the importance of community, and why he believes the traditional Google search experience won’t survive the next five years.

From the PPC Wild West to a career in search

Finn’s introduction to digital marketing came around 2004 and 2005 while working at a technology company.

Initially working in sales support, he began shadowing the company’s digital marketer and taking on additional work, learning platforms such as Overture and Yahoo before Google AdWords became dominant.

“I just found that I loved it and I was good at it,” Finn said.

His experience wasn’t limited to paid search. Finn also worked heavily in SEO, including an in-house role where he was tasked with ranking for highly competitive terms such as “student loans.”

That combination gave him a front-row seat to an era when marketers had substantially more granular control over both paid and organic search.

PPC used to reward precision

Finn believes PPC was easier for highly skilled practitioners in its early years because there was a much wider gap between advertisers who knew what they were doing and those who didn’t.

Advertisers could tightly control keywords, ads and landing pages. Exact match behaved much more like its name suggested, allowing marketers to construct extremely specific experiences around individual searches.

Teams could obsessively test everything from ad copy to tiny landing-page design changes.

“If you were good, it was easier,” Finn said.

Today, automation has narrowed some of that expertise gap. Google can make less sophisticated advertisers more competitive, while products such as Performance Max and AI Max can also cause advertisers to compete in auctions they may not have deliberately selected.

The moment PPC proved itself

One of Finn’s earliest “this actually works” moments came from helping a local costume shop in Rochester, New York.

The business had a single physical store, but paid search allowed it to compete for customers far beyond its local market.

Seeing how much revenue a small business could generate from digital advertising helped convince Finn of paid search’s potential.

“If you have somebody that’s good, you have somebody that cares about your money and is trying to do the best for you, it was awesome just to see,” he said.

Search Engine Land became part of the education

Finn vividly remembers Search Engine Land launching.

At the time, industry education was much harder to find. He listened to WebmasterRadio throughout the working day because he was essentially his own department and needed a way to learn from other search marketers.

When Danny Sullivan announced he was leaving Search Engine Watch to launch Search Engine Land, Finn remembers it feeling like major industry news.

Search Engine Land, its Daily SearchCap and eventually SMX became important sources of both information and community.

That eventually came full circle. Finn went from consuming Search Engine Land’s coverage to contributing more than 200 articles himself.

The change that signaled automation was taking over

For Finn, one of PPC’s biggest turning points arrived in 2014 when Google required campaigns to use close variants for exact match.

He covered the change for Search Engine Land at the time.

Until then, skilled advertisers could exercise extremely precise control over matching, negatives, ads and landing pages.

Once “exact” stopped necessarily meaning exact, Finn saw where Google Ads was heading.

“That was the beginning of automation,” he said.

Expanded text ads were another major milestone, but the changing definition of exact match represented something bigger: advertisers were beginning to surrender some of the control that had defined paid search.

His takeaway from the years since is simple: marketers can object to automation, but eventually they have to learn how to succeed within it.

Broad match went from something to avoid to something Finn uses everywhere

Few examples illustrate PPC’s transformation better than broad match.

Finn once helped write a book effectively warning advertisers not to use it. Today, he uses broad match across his accounts.

The reason is that broad match itself has changed.

Modern matching can consider signals beyond the literal keyword, including previous searches and landing-page context. Combined with Smart Bidding, Finn believes advertisers should increasingly evaluate the outcome rather than obsess over whether every individual search term appears perfect.

“You’re not doing keywords, you’re doing intent,” he said.

There are still obvious negatives advertisers should manage. But if a query looks unusual while the campaign is generating profitable results, marketers need to consider that Google may be using context they cannot see from that individual search alone.

Smart Bidding may be PPC’s biggest innovation

Asked to identify the best PPC innovation of the past 20 years, Finn eventually landed on Smart Bidding.

Its value is partly the removal of tedious manual bid management.

Instead of repeatedly adjusting individual keywords, advertisers can communicate the business result they want — such as a target CPA or target ROAS — and allow the system to determine auction-level bids.

That doesn’t mean blindly trusting Google.

Finn remains conscious that advertisers and their competitors are effectively telling the same advertising platform what they are willing to pay. But when the underlying measurement is sound, he sees automated bidding as a major improvement over manually managing thousands of bids.

The old Google felt more advertiser-first

One thing Finn misses from the early search era is Google’s relationship with advertisers.

Google was the challenger. Teams were smaller, representatives were more consistent and advertisers could interact much more directly with the people building the platform.

As Google matured into a massive public company, Finn believes that relationship inevitably became more corporate.

Changes ranging from advertiser fees to payment policies illustrate the tension between what’s best for advertisers and what’s best for Google’s business.

Finn understands why that happens — Google ultimately has shareholders to answer to — but the relationship feels very different from the early AdWords era.

Brand bidding remains a necessary evil

Finn also misses the period when Google was much stricter about advertisers bidding on competitors’ brand terms.

Today, competitive bidding — whether deliberate or emerging through automated campaign types — means many advertisers feel they need to spend money protecting searches for their own brands.

That’s something Finn fundamentally dislikes.

He cautioned against agencies celebrating huge amounts of revenue attributed to branded search without acknowledging that those customers were already looking for the company.

Yet the competitive environment means advertisers can’t simply pretend brand bidding doesn’t exist. They have to respond to the marketplace as it actually operates.

Mistakes still matter in an automated world

Finn also recalled some painful PPC mistakes.

Early in his career, a campaign intended to spend roughly $30,000 in a month was accidentally configured around that figure as a daily budget.

Years later, an issue involving the transfer of Google campaign data into Microsoft Advertising resulted in ads and keywords from one client appearing in another client’s account. His agency reimbursed the affected client for the incorrect spend.

The experiences reinforced the importance of checking automated imports, account structures and campaign settings carefully.

Automation may remove manual work, but it doesn’t remove accountability.

PPC community became one of the industry’s biggest strengths

The search community Finn remembers from early conferences looked very different.

SEO practitioners seemed like the industry’s rock stars, while PPC sessions could feel comparatively corporate and buttoned-up.

That distinction has faded.

Finn believes the PPC community has become particularly strong at sharing information, discovering changes and helping less-visible practitioners gain recognition.

That’s also something he tries to do through Marketing O’Clock: highlight people doing strong work even when they aren’t established industry personalities.

For Finn, helping good practitioners and agencies succeed ultimately benefits the whole industry.

AI Mode is the future — whether marketers like it or not

Finn has a stark prediction about the future of Google Search.

Marketers don’t get to decide whether AI search happens.

“You don’t have any say. You can whine all you want. It’s the future for sure,” he said.

Finn sees AI Overviews as a transitional product and AI Mode as a much clearer indication of where Google is ultimately heading.

His prediction: the familiar Google experience built around traditional blue links won’t exist in its current form five years from now.

That doesn’t mean links can disappear completely. Google and other AI platforms still depend on an ecosystem of publishers and websites producing information.

The challenge will be finding an economic model that keeps that ecosystem sustainable while search interfaces increasingly answer questions themselves.

Authenticity becomes more valuable as AI content grows

If AI makes competent content cheap and ubiquitous, Finn believes human authenticity becomes a bigger differentiator.

People can already recognize formulaic AI-generated LinkedIn posts and other content, he said. The question for marketers therefore isn’t whether AI can produce something “good.”

It’s how humans produce something great enough to stand out from an endless supply of good-enough machine-generated material.

That means having a recognizable voice, expressing genuine opinions and creating things audiences deliberately want to consume.

The same principle applies inside paid media. Complex businesses still require marketers to understand different conversion types, business economics, measurement and how all those signals should inform bidding.

AI can accelerate execution. Finn doesn’t believe it replaces the thinking behind it.

The next generation should stop waiting and start creating

Asked what advice he’d give his younger self, Finn’s answer was confidence — and action.

If you want to write, start writing. If you want a podcast, make one. If you want to participate in the industry, reach out to people.

Don’t spend forever perfecting the logo or planning every possible outcome.

Some projects will fail, Finn said, but failure provides information that makes the next attempt better.

That philosophy mirrors his own route into Search Engine Land. When a colleague no longer had enough time for his column, Finn volunteered to take it over. That opportunity eventually became hundreds of contributions across Search Engine Land and its sister publications.

Building people is the career achievement that matters most

Despite decades of campaigns, articles, speaking appearances and industry work, Finn said his proudest accomplishment is Cypress North and the people the agency has developed.

Operating in Western New York, where the digital marketing talent pool can be smaller than in major advertising hubs, the agency has invested heavily in finding people with potential and teaching them the craft.

Some joined without digital marketing experience and developed into skilled practitioners.

For Finn, watching those people improve is ultimately more meaningful than any individual campaign result.

The bottom line

PPC has moved from an era where practitioners could obsess over individual keywords, bids and landing pages to one where algorithms increasingly make those decisions.

Finn’s career has spanned that entire transition.

His message isn’t to fight the direction platforms are moving. It’s to understand where human expertise still creates an advantage: business judgment, measurement, creativity, authenticity and knowing when automation is actually delivering the outcome that matters.

The tools have changed dramatically over 20 years. The responsibility to make them work for the business hasn’t.

Looking to take the next step in your search marketing career?

Below, you will find the latest SEO, PPC, and digital marketing jobs at brands and agencies. We also include positions from previous weeks that are still open.

Newest SEO Jobs

(Provided to Search Engine Land by SEOjobs.com)

• Link Outreach and Directory Submissions
  August 26, 2026
  Company: FoundCiteSalary: USD 15-20 per hour, 5 to 10 hours per weekLocation: Remote WorldwideDate posted: August 26, 2026FoundCite measures how often AI assistants name a brand when buyers ask for a recommendation, then publishes the content that closes the gap. It is a new product with a real problem: zero referring domains and content that deserves to be found. I […]
• SEO Automation
  August 26, 2026
  Company: EnebaLocation: Remote, United StatesDate posted: August 26, 2026About Your Team Join our Marketing Team, where we're sustainably driving growth with goals to become the leading marketplace for gamers. We obsess over merging data-driven optimization, entertainment, and creativity, and our SEO team is where this comes to life on the product side. You'll join the onsite side of SEO – […]
• Associate Director, SEO Specialists
  August 25, 2026
  Company: OuterBoxSalary: $85,000-$105,000 per year USDLocation: Remote, United StatesDate posted: August 25, 2026The Associate Director, SEO Specialist owns the day-to-day operations of the SEO Specialist team, acting as the primary point of leadership for the Specialists and Leads who execute technical and tactical SEO work across the department. This role manages the intake and prioritization of day-to-day SEO tasks coming […]
• SEO Specialist
  August 25, 2026
  Company: AmplifyedSalary: 80,000Location: Remote USDate posted: August 25, 2026 Do you love nerding out on SEO and working with clients? If your friends & family are sick of hearing about your latest search rankings, then we’re your kind of people and you will love this job. $80k in year 1 + potential bonuses You will get an absolute masterclass on […]
• Digital Marketing Strategist (Account Management, SEO, PPC, Social)
  August 25, 2026
  Company: Volume NineSalary: $55,000-$70,000 per year USDLocation: Remote, United StatesDate posted: August 25, 2026Keynotes: 100% Remote Role. US-Based candidates only. Salary Level: $55,000 – $70,000/year (depends on experience) At Volume Nine, we help companies create a predictable, steady stream of customers using smart, modern digital marketing. The industry changes fast (you’ve probably noticed ), so our mission is simple: stay […]
• Growth Specialist SEO Mdr. / Sr (€3.250 en €4.500)
  August 25, 2026
  Company: GradientLocation: Remote, United StatesDate posted: August 25, 2026Over Gradient Gradient helpt B2B-bedrijven aan voorspelbare, ontworpen groei in plaats van losse campagnes en kortetermijntactieken. Wij brengen marketing, sales en service samen in één geïntegreerd Growth System, met strategie in het middelpunt en performance, content, RevOps en media die op elkaar zijn afgestemd, vaak aangedreven door HubSpot. Een senior-led team werkt […]
• SEO Strategist
  August 24, 2026
  Company: Noon EcommerceSalary: 60K–80K a yearLocation: Remote, United StatesDate posted: August 24, 2026Remote-Friendly SEO Strategist for Growth & Local SEO in Pleasant Ridge at Noon Ecommerce | Apply now! At a Glance Tasks: Boost clients’ search visibility and traffic through expert SEO strategies. Company: Dynamic ecommerce agency with a focus on growth and collaboration. Benefits: Competitive salary, health insurance, paid […]
• SEO/GEO Strategist
  August 24, 2026
  Company: Winston Digital MarketingSalary: $70,000 per year USDLocation: Remote, United StatesDate posted: August 24, 2026Senior SEO Strategist (70K/year – Full Time) The Role You’ll own SEO strategy and execution end-to-end for major cannabis brands, finance clients, and ambitious local businesses. Ship audits. Build roadmaps. Implement fixes. Measure wins. Do it all with complete autonomy and zero micromanagement. We’re a small, […]
• Senior SEO Manager
  August 24, 2026
  Company: Single GrainSalary: $85,000-$100,000 per year USDLocation: Remote, United StatesDate posted: August 24, 2026Senior SEO Manager job at Single Grain Remote. We’re looking for a humble, hungry, and smart Senior SEO Manager to join our team. In this role, you will be responsible for end‑to‑end management of multiple client accounts, driving performance through strategic campaign optimization while building strong client […]
• Senior SEO Account Manager
  August 23, 2026
  Company: Profile 29Salary: £54K–£66K a yearLocation: Remote, United StatesDate posted: August 23, 2026Senior SEO Account Manager Salary to £60k Benefits include medical insurance, pension + more Role is fully remote (UK) About Them: This company is a multi-award winning, highly reputable digital marketing agency. They are built on hard-work, creativity and laughter, and every member of their team brings an […]

Newest PPC and paid media jobs

(Provided to Search Engine Land by PPCjobs.com)

• Remote Demand Generation + Growth Marketing Manager
  August 28, 2026
  Company: grabjobsSalary: $100,000 – $125,000Location: United States (residency required)Date posted: August 28, 2026 About Barti Barti is a venture-backed startup on a mission to build the future of eye care. We're building groundbreaking, AI-powered software that transforms how practices document care, run their operations, and serve patients. Our goal is to craft exceptional user experiences that let eye care providers […]
• Remote Growth Marketing Manager, Buddy
  August 28, 2026
  Company: grabjobsSalary: $110k – 170kLocation: United States (residency required)Date posted: August 28, 2026 Description Skylight is a technology startup building the OS of the family. We make Skylight Calendar, the smart calendar loved by millions of families (plus Wired and the Wirecutter ). Our latest product is Calendar 2 , which just launched to rave reviews. Our mission is to […]
• Remote Growth Marketing Manager, Calendar
  August 28, 2026
  Company: grabjobsSalary: $110k – 150kLocation: United States (residency required)Date posted: August 28, 2026 Skylight is a technology startup based out of Los Angeles and San Francisco. Our mission is to connect loved ones by creating the world’s simplest products and services that improve family life. We make consumer products, like Skylight Frame and Skylight Calendar, that are loved by millions […]
• Remote Sr. Manager, PPC Marketing
  August 28, 2026
  Company: grabjobsSalary: $125,000 – $135,000Location: United States (residency required)Date posted: August 28, 2026 Job description About Certus We sit at the intersection of people and technology—building training and education solutions that help more than 4 million learners each year enter, advance, or reinvent their careers. Our customers are some of the world's most recognized brands: Amazon, Siemens, Geico, Chick-fil-A, and […]
• Remote Advertising Paid Search Manager – Ecommerce
  August 28, 2026
  Company: grabjobsSalary: $65,000-$88,000Location: United States (residency required)Date posted: August 28, 2026 Rithum is the world’s most trusted commerce network, accelerating how brands, suppliers, and retailers work together to deliver seamless e-commerce experiences. We provide an unmatched platform for brands and retailers, enabling them to accelerate growth, optimize operations across channels, scale product offerings and enhance margins. Today, more than 40,000 […]

Other roles you may be interested in

Paid Media Manager (Search + Social), Creative Circle (Coppell, TX)

• Salary: $80,000 – $100,000
• Own and execute the paid digital strategy across Meta, TikTok, YouTube, Google Paid Search, retail media, and more – building the plan, launching it, and managing it day to day.
• Monitor and optimize campaigns daily and weekly against core KPIs (ROAS, conversions, traffic, engagement, sales); analyzing performance and translating data into actionable recommendations.

SEO Manager, M+C Saatchi Group (Hybrid, Washington, DC)

• Salary: $67,500 – $72,500
• Develop and execute enterprise SEO strategies that improve organic visibility, search rankings, and qualified traffic across websites and digital properties.
• Conduct comprehensive keyword research, search intent analysis, and competitive assessments to identify opportunities for content growth and improved search performance.

SEO and GEO Manager Halstead Media (Remote)

• Salary: $62,000 – $68,000
• Own SEO and GEO strategy and performance across an assigned portfolio of clients
• Build annual, quarterly, and monthly SEO priorities based on client goals, opportunities, and performance

Paid Media Manager (SEM/PPC), YPM (Irvine, CA)

• Salary: $95,000 – $100,000
• Build, launch, manage and optimize across multiple paid digital channels / search engine platforms media plans (Google, Bing, etc.) in varying verticals.
• Develop and maintain strong, key relationships with clients/brands; communicate with clients and respective Google reps on strategic direction of PPC campaign initiatives, identifying key findings and recommendations.

Demand Generation Manager, Seer Interactive (Remote)

• Salary: $90,000 – $105,000
• Campaigns: Build and run demand campaigns end to end, sequencing what the team produces across email, social, events, comms, paid, and outbound rather than letting it ship one-off
• Programs: Own our newsletters end to end: strategy, curation, production, reporting. Own the Signal webinar series and our event and conference presence, from promotion through logistics through follow-up

Senior SEO Strategist, elk Marketing (Remote)

• Salary: $95,000 – $115,000
• Lead keyword research, site audits, content strategy, and campaign planning across your assigned clients
• Develop multi-month strategic plans that demonstrate exceptional foresight & prioritization, ensuring elk & client teams are aligned and executing strategic implementation effectively & efficiently

Sr. Manager Paid Search, Ninja Trader (Hybrid, Chicago, IL)

• Salary: $115,000 – $165,000
• Own paid search strategy across Google Ads, Bing, and other engines, managing seven figures in monthly media spend
• Build and scale a non-brand experimentation program, designing structured tests to unlock incremental, profitable growth

Marketing – Manager, Paid Media and Activation, DeVry University (Remote)

• Salary: $95,000 – $105,000
• Lead activation and operational management of paid media campaigns across digital channels (social, programmatic, video, display, etc.)
• Partner with Senior Manager to execute media plans aligned to university goals and enrollment priorities

Digital Marketing Manager, Paid Social & PPC, 80Twenty (Hybrid, Newark, DE)

• Salary: $90,000 – $110,000
• Own paid media strategy and execution across Meta Ads, LinkedIn Ads, TikTok Ads, Google Ads, Demand-Side Platforms (DSPs) and other media platforms
• Drive the growth of product lead volume while operating within defined budgets, cost-per-acquisition (CPA) and return on ad spend (ROAS) targets

Paid Media Manager, Clients Blackbox, Inc. (Remote)

• Salary: $90,000 – $125,000
• Build, launch, and optimize Meta advertising campaigns focused on lead generation and appointment booking
• Manage campaign budgets, audience targeting, bid strategies, and creative testing

Note: We update this post weekly. So make sure to bookmark this page and check back.

]]> https://searchengineland.com/google-says-better-youtube-creative-can-more-than-double-roi-486180 Fri, 28 Aug 2026 15:27:20 +0000 https://searchengineland.com/?p=486180 Google’s new YouTube creative guidance highlights the ad elements associated with stronger conversions and ROI.]]>

Google has published a new Creative guidance on YouTube help document, giving advertisers specific recommendations for designing ads that fit the platform — and sharing new data on the creative elements associated with higher conversions.

The guidance covers everything from using human voices and showing brands early in videos to removing fake buttons from images and making creative look more native to YouTube.

Why it matters. Google is putting numbers behind the importance of creative quality. The new document cites Nielsen research finding that creative accounts for 49% of campaign ROI, while Ekimetrics research found that improving creative can more than double ROI on YouTube.

That makes creative optimization potentially as important as many of the targeting and bidding decisions advertisers spend their time adjusting.

What the data says. Google shared three findings specifically for Demand Gen in-stream skippable ads, based on U.S. data from July 2026.

Including a human voice was associated with 12% higher conversions on average, adding supers or a text overlay with 3% higher conversions, and visualizing the brand within the first five seconds with 4% higher conversions.

Human voices stand out. Of those findings, voice had the largest reported conversion lift.

Google’s guidance suggests advertisers shouldn’t treat audio as an afterthought when producing YouTube creative. A human voice can be an important part of designing the ad for the environment in which it will actually be consumed.

Don’t fake the interface. For image creative, Google recommends removing embedded buttons, arrows and other elements that appear clickable but aren’t.

Google says these “baked-in” interface elements can compete with YouTube’s actual calls to action, adding cognitive load and potentially confusing users rather than directing them toward the advertiser’s site.

Make ads look like they belong. Google is also encouraging advertisers to design images that blend naturally into YouTube feeds.

That means using full-bleed, high-resolution photography or lifestyle imagery while avoiding oversized logos, empty solid areas and black bars that can make an ad look fragmented or unnatural within the platform.

Put people in the creative. Google says images showing a visible human presence consistently outperform other assets and recommends featuring faces or people interacting with products when appropriate for the campaign.

The broader message is that YouTube creative should lean into the human and visual nature of the platform rather than simply repurposing assets designed for other advertising environments.

Why we care. As Google increasingly automates bidding, targeting and campaign management, creative remains one of the biggest areas advertisers can directly influence. Google’s new guidance also provides some useful benchmarks for testing — particularly the 12% conversion difference associated with human voice — rather than simply telling advertisers to “make better creative.”

The bigger picture. Google’s recommendations reinforce that simply having enough assets isn’t the same as having assets designed for YouTube. Advertisers running Demand Gen and other YouTube campaigns may want to test not only different messages, but how those messages are presented — including voice, branding, text overlays, imagery and how naturally the ad fits into the YouTube experience.

Bottom line. Google’s new help document gives advertisers a clearer creative playbook for YouTube: use human voices, establish the brand early, consider text overlays, avoid fake UI elements and create imagery that feels native to the platform.

Google is making three unversioned changes to the Display & Video 360 API and Structured Data Files in October, including removing some content exclusion options and introducing new requirements for YouTube responsive ads.

Because the changes are unversioned, developers with existing integrations may need to make adjustments rather than waiting until they migrate to a newer API version.

What’s changing. Starting Oct. 1st, advertisers will no longer be able to use targeting to exclude specific digital content labels through the Display & Video 360 API.

The change affects available TARGETING_TYPE_DIGITAL_CONTENT_LABEL_EXCLUSION targeting options as well as valid values in the Digital Content Labels - Exclude column within Line Item Structured Data Files.

Sensitive category exclusions. Google is also removing the ability to exclude the majority of sensitive categories using targeting from Oct. 1.

That change affects TARGETING_TYPE_SENSITIVE_CATEGORY_EXCLUSION options in the API and the Brand Safety Sensitivity Setting and Brand Safety Custom Settings columns in Line Item Structured Data Files.

YouTube ads get new requirements. A separate change takes effect Oct. 12, when Google will require a business name and logo when developers create or update YouTube responsive ads. The requirement applies when default business name and logo values haven’t already been assigned to the parent advertiser.

It will also affect advertisers and developers uploading ads through Ad Structured Data Files.

Why we care. The removal of content label and sensitive category exclusion options could require advertisers and platforms relying on those targeting controls to review their existing brand-safety workflows. Meanwhile, integrations that create or update YouTube responsive ads will need to ensure the required business identity assets are available before Oct. 12 to avoid problems with their workflows.

What developers should do. Google is advising developers who believe their integrations will be affected to follow the recommended actions outlined within its individual deprecation descriptions.

That makes it worth auditing integrations ahead of October, particularly those that manage content exclusions or automate YouTube responsive ad creation through the API or Structured Data Files.

Key dates. The targeting changes take effect Oct. 1, removing specific digital content label exclusions and most sensitive category exclusions. The new business name and logo requirement for YouTube responsive ads follows on Oct. 12.

Bottom line. Display & Video 360 integrations have two October deadlines to prepare for, with Google reducing some targeting-based exclusion controls while adding mandatory business identity requirements for YouTube responsive ads.

There’s a familiar conversation in almost every marketing organization: “We need more leads.” Then, a few weeks later: “The leads we’re getting aren’t good enough.”

So the marketing team tightens targeting, changes creative, adds qualification requirements, and raises the bar for conversion. Lead volume drops, and the cycle begins again.

The problem is the assumption that lead quantity and lead quality are competing strategies. They’re different growth objectives that require different inputs, signals, and channel levers.

The question you should be asking is: “What does the business need right now, and have we given our advertising platforms the signals necessary to achieve it?”

Start with the business objective, not the media metric

The first mistake is jumping directly into channel metrics. If the business says it needs 30% more leads, the media team shouldn’t automatically start optimizing toward cost per lead.

If the business says it needs more revenue from marketing, lowering CPL might actually make the problem worse. A $30 lead that never becomes a customer isn’t necessarily better than a $100 lead that consistently closes.

Think about the funnel as a progression: Impression → Click → Lead → Qualified lead → Opportunity → Customer. Your optimization strategy should reflect where the business is trying to create growth.

When volume is the priority

A volume objective makes sense when:

• You have insufficient leads to support sales capacity.
• A new campaign or market needs to build demand.
• Your sales team can handle substantially more opportunities.
• You’re intentionally expanding your addressable audience.
• You need more data before you can confidently identify quality signals.

In this scenario, you want to remove unnecessary constraints that prevent the platform from finding more potential customers.

When quality is the priority

A quality objective makes more sense when:

• Sales is overwhelmed with poor-fit leads.
• Lead-to-opportunity rates are weak.
• Close rates are declining.
• Customer acquisition costs are rising despite healthy lead volume.
• Sales capacity is limited.
• You have enough CRM data to identify what a valuable lead looks like.

You want to find more people who resemble those who ultimately become customers.

Dig deeper: 5 ways to improve PPC lead quality

If you need more leads, remove friction and expand the funnel

When volume is the priority, look for opportunities to increase eligible audience size and reduce barriers to conversion.

That might mean:

• Expanding keyword coverage.
• Broadening match types.
• Testing additional geographic markets.
• Expanding paid social audiences.
• Using automated audience expansion.
• Introducing additional inventory, such as Display, Demand Gen, YouTube, or other video placements.
• Testing shorter forms.
• Reducing unnecessary qualifying questions.
• Creating additional landing pages or offers.
• Increasing creative variety.

Google Search

For volume, look beyond the handful of exact queries that already perform well.

Expand keyword coverage and give Google’s automation more room to find incremental demand. Broad match and automated bidding can help when conversion tracking and negative keyword management are strong enough to protect against obvious waste.

You can also expand beyond Search into Google’s broader inventory when incremental volume is the goal. But there’s an important caveat: More inventory means more responsibility for your conversion signals.

If Google is rewarded for every form submission, it has little reason to distinguish between a future customer and someone who submitted a form simply because it was easy.

Paid social

Paid social can also be useful when reach and lead volume are the priority. The volume-oriented playbook generally means giving the algorithm more room to work:

• Broader audiences.
• Automated placements.
• More creative variations.
• Fewer targeting restrictions.
• Lower-friction lead forms.
• More conversion opportunities.

The objective is to maximize the number of people who can potentially enter the funnel without unnecessarily restricting the algorithm.

That doesn’t mean “target everyone.” It means being intentional about which constraints are actually helping performance and which are simply limiting scale.

Dig deeper: How to measure paid social’s impact on paid search performance

If quality is the priority, teach the platforms what ‘good’”’ means

This is where many lead-generation programs break down.

The advertising platform sees 1,000 form submissions. Your CRM sees 1,000 leads → 300 qualified → 75 opportunities → 20 customers.

If the platform only receives the first data point, it doesn’t know what happened to the other 999 people.

You’re asking an increasingly sophisticated machine-learning system to optimize with an incomplete definition of success. The solution is to close the feedback loop.

Google’s current lead generation framework supports qualified-lead and converted-lead conversion goals, allowing advertisers to feed deeper-funnel outcomes back into Google Ads. Google says these goals can provide more information for optimization and help advertisers balance lead volume and quality.

Meta’s Conversions API similarly connects CRM and other first-party data to its optimization systems. Meta says advertisers can use down-funnel lead data to help optimize for people more likely to become customers rather than simply maximizing lead submissions.

In other words, the conversion doesn’t necessarily have to be the form fill.

It can be what happens after the form fill.

Dig deeper: How to improve Google Ads lead quality with CRM data

Google Search

When quality is the priority, consider optimizing toward a deeper-funnel event:

• Qualified lead.
• Sales-qualified lead.
• Opportunity.
• Application.
• Purchase.
• Closed customer.

The challenge is finding the lowest-funnel event with enough volume to provide a useful signal. If you generate only a handful of customers each month, optimizing exclusively toward customers may not provide enough data for effective automated bidding.

A higher-volume but still meaningful event — such as qualified leads or opportunities — may be the better optimization point. Choose the deepest conversion that still gives the platform enough signal to learn.

Paid social

The same principle applies to Meta and other paid social platforms.

Connect CRM data where possible and use your actual business outcomes to inform optimization. Meta specifically supports sending down-funnel CRM data through the Conversions API so its systems can learn which leads are more likely to convert.

Help the platform find people who are more likely to become customers, rather than simply people who submit forms. 

Don’t overlook creative: It can be a quality filter

Creative is one of the most underappreciated levers in lead quality. Your messaging determines who recognizes themselves in your offer.

An ad that simply says “Get started today” is likely to attract a broader audience than one that communicates something specific about the product, price, experience, eligibility, or outcome.

That isn’t necessarily bad. But creative can function as a qualifier. When quality is the priority, make your value proposition more explicit. Tell people:

• Who the product is for.
• What problem it solves.
• What makes it different.
• What the commitment looks like.
• What the next step requires.

You may see CTR or raw lead volume decline. That’s OK if the people who remain are more likely to progress through the funnel.

When volume is the priority, the opposite may be appropriate. Test broader messages, more entry points, and more creative concepts designed to capture different motivations.

Your business objective should influence your creative brief.

Different channels can have different jobs

Another common mistake is expecting every channel to chase the same KPI. They don’t have to.

• Search can capture existing intent. 
• Paid social can create or shape demand.
• Display and video can expand reach.
• Retargeting can bring high-intent prospects back into the funnel.

For example:

The point isn’t that one channel is inherently better for quality or quantity. It’s that different channels can play different roles in the customer journey.

Your media plan should reflect those jobs rather than forcing every channel to chase the same CPL.

The most dangerous KPI is the one that looks good in isolation

A $25 CPL looks great. Until you discover those leads convert to customers at 1%. A $100 CPL looks terrible. Until you discover those leads convert at 20%.

That’s why you should monitor multiple layers of performance simultaneously.

• Volume metrics
  • Leads.
  • CPL.
  • Conversion rate.

• Quality metrics
  • Qualified lead rate.
  • Cost per qualified lead.
  • Lead-to-opportunity rate.
  • Opportunity-to-customer rate.

• Business metrics
  • Customer acquisition cost.
  • Revenue.
  • Lifetime value.
  • Return on ad spend.

The exact metrics will vary by business, but the principle is universal:

Don’t let the easiest metric to measure become the definition of success.

The answer isn’t always quality or quantity

Sometimes the business needs both. You may have a highly efficient campaign generating 50 great leads, but the business needs 100. Instead of sacrificing quality to chase volume, establish a quality floor and a volume ceiling.

• Define what makes a lead valuable: Work with sales and leadership to identify the characteristics and actions that correlate with meaningful business outcomes.
• Feed those signals back into your platforms: Connect CRM and offline conversion data where possible.
• Establish a minimum acceptable quality threshold: Know the lowest qualified-lead rate or the highest cost per qualified lead the business can tolerate.
• Scale within that boundary: Expand audiences, inventory, keywords, and creative to find incremental volume.
• Watch what happens to quality as you scale: If quality deteriorates rapidly, identify which lever is responsible.
• Pull back selectively: Don’t necessarily shut down the entire campaign. Reduce the specific source of inefficient volume.

Stop asking, ‘How do we get more leads?’

The better question is: “What kind of growth does the business need?”

If the answer is volume, give your channels room to find more people. If it’s quality, give them better signals. If it’s revenue, connect your advertising platforms to the outcomes that actually generate revenue.

If the business needs both, give different campaigns and channels clearly defined jobs, and make sure the metrics you’re optimizing toward reflect those jobs.

The real advantage is building a marketing engine that can scale when the business needs volume, tighten when quality starts to suffer, and continuously learn from what happens after the lead is generated.

]]> https://searchengineland.com/use-claude-for-seo-dont-let-claude-do-seo-485931 Fri, 28 Aug 2026 14:00:00 +0000 https://searchengineland.com/?p=485931 Claude can accelerate research, analysis, and drafting, but SEO execution still needs human judgment to avoid costly mistakes.]]>

Claude is one of the best research assistants an SEO can have. Give it a keyword, and it can map intent, spot content gaps, and draft outlines in minutes. But give it write access to your site, and it can take the fastest path to a plausible-looking answer.

I’ve learned to keep a human between AI and any live page. Claude is genuinely useful for SEO research, analysis, and drafting, but it can be quietly wrong when it’s left to execute on its own.

AI can process information and produce recommendations quickly. SEO execution requires judgment about whether those recommendations are actually right for a specific site, page, and search intent. Keep AI in the research and drafting seat, and keep a human responsible for what goes live.

What Claude is actually good at in SEO work

Claude is helpful in the research and analysis phase of SEO, before anything touches a live page. Ask it to cluster a list of keywords by intent, summarize what the top-ranking pages for a query have in common, or draft a first-pass content outline, and it’s fast and useful.

It reads a page and tells you what’s thin.

It reads a competitor’s article and tells you what angle it’s missing.

It turns a messy spreadsheet of search terms into a structured content plan in a few minutes of back-and-forth.

That’s real value, and it shows in the adoption numbers. In Keyword.com’s State of AI in SEO 2026 survey (97 usable responses, skewed toward lean teams and service providers), 87% said they use AI regularly or as a core part of how they deliver SEO work, and 78% of respondents said they use Claude, ahead of ChatGPT at 57%.

The tools are being used. The question is what they’re being used for.

Research, synthesis, and first drafts are good jobs. Publishing decisions aren’t, for a practical reason: The two tasks require different things from the model.

Research needs a model that can hold a lot of context and generate plausible-sounding options fast.

Execution needs a model that knows when a plausible-sounding option is actually wrong for this specific page, site architecture, and keyword map. Claude, like every current AI model, doesn’t reliably know that.

Dig deeper: How to build a Claude Code-powered second brain for agency work

Why ‘looks plausible’ isn’t the same as ‘is correct’ for SEO

One failure should worry anyone handing Claude the keys to a CMS. When an AI model is asked to solve a content problem and doesn’t have a new answer, it can produce something that resembles a correct answer instead of admitting it doesn’t have one.

In coding, that shows up as invented function names. In SEO, it shows up as duplicated pages with new titles.

I used Claude to review Google Search Console, recommend target keywords for our own AI Website Grader, and build the pages those keywords needed. Instead of writing purpose-built content, Claude cloned the homepage into two new URLs, /seo-grader and /content-grader, changed the title tag and H1 to match the new keywords, and reused most of the homepage’s body copy.

On paper, each “page” now targeted a new term. In practice, it was the homepage’s content living at three URLs, competing with itself, exactly the kind of cannibalization a competent SEO would never do.

I left those two pages live on purpose, as a running illustration. Each still lifts most of its body copy from the homepage, and six months of Google Search Console data show the result. The dedicated /seo-grader and /content-grader pages have earned zero impressions and zero clicks.

Every query they were built to win goes to the homepage instead: “content grader” (487 impressions, average position 9.2), “seo grader” (position 10.6), and “ai content grader” (position 5.3).

The clone didn’t just risk cannibalization. It generated no impressions or clicks, while the homepage sits stuck at the bottom of Page 1 for the exact terms a dedicated page should own.

Microsoft has confirmed the same dynamic on the AI side: Bing’s models group near-duplicate URLs into a single cluster and may pick an unintended page as the representative source, so a clone can even hand AI answers the wrong URL.

I’ve seen SEOs make plenty of mistakes over the years. I have never seen one deliberately duplicate a ranking page and slap a new title on it, because every SEO knows it’s a self-inflicted wound.

Claude did it anyway, with total confidence, no caveat, and no flag that the “new” page was actually a clone.

The same mistake happened twice

If this had happened once, I’d chalk it up to a bad prompt on my part. It happened again, on a completely different site, run by someone else.

My son Caleb started a project called ScryPrice, a price-comparison engine for Magic: The Gathering cards. He asked Claude for SEO keyword recommendations, the same kind of request I’d made. Claude’s response was to generate a batch of new pages built to target those keywords.

When we checked them, every one of those “new” pages was a copy of the homepage with nothing changed except the title tag. No unique product context, no differentiated content, no reason for a search engine to treat any of them as distinct from the page they were cloned from.

Same failure mode, same root cause, two unrelated projects, months apart, with no shared prompt or workflow between us.

That’s what makes this worth writing about instead of shrugging off: It’s what can happen when an AI model is given execution authority over a content or SEO task and no guardrail forces it to actually build something new instead of reshaping something that already exists.

Dig deeper: How to train Claude to sound like your brand

Other SEO failure modes show the same pattern

There’s a crawlability version of the same failure. An SEO developer who posts as @rentierdigital measured it: 

• “[C]laudebot downloads your [JS] bundle in 24% of its requests and never executes it. it cannot read the thing it helped you build.”

When an AI agent builds a JavaScript-heavy page that depends on client-side rendering, the page can look finished while remaining difficult for crawlers to read and index.

Other SEO practitioners have described similar problems with AI-generated pages. SEO consultant Robert May put it plainly: 

• “The problem’s not AI, it’s using AI to create hundreds of pages that should never have existed in the first place.”

Digital marketer Scott DeSapio identified another problem with using AI to scale content: 

• “One page trying to rank for five different searches usually ranks for none. Each URL should serve one clear search intent.”

An account posting as @digispot_ai listed a similar set of symptoms: 

• “Missing internal links. Duplicate intent. Wrong schema. Cannibalized keywords. Pages Google never indexes.”

None of these are the same mistake I saw, but they’re all variations of the same problem: Give an AI model autonomy over an SEO task, and it can optimize for producing something that looks finished rather than producing something that’s actually right.

The highest-stakes problem isn’t even SEO

The same pattern shows up far beyond SEO when AI agents get real execution authority without a human checkpoint.

In July 2025, SaaS founder Jason Lemkin said Replit’s AI coding agent deleted the company’s production database despite instructions to freeze code and data changes, wiping records for more than 1,200 executives and more than 1,190 companies. Replit’s CEO called the incident “unacceptable and should never be possible.”

That’s not an SEO story, and I’m not claiming it is. The point is the failure pattern: An AI agent given execution authority, operating without a human checkpoint, can optimize for producing an output rather than producing the right output.

The SEO version is quieter. Nobody’s database gets wiped when Claude clones a page instead of writing one. But the underlying problem is the same: The agent can produce something that looks finished without recognizing that it’s wrong.

Give AI the research, keep humans on execution

None of this means AI models are broken or useless for SEO. The job just needs to be split in two, and only one half belongs to the AI.

Give Claude the research and analysis work

Keyword clustering, search intent classification, content gap analysis against competitors, technical audit summaries, first-draft outlines, and copy.

This is where Claude’s ability to process a lot of information quickly and surface patterns does the most good. It cuts hours out of the front end of SEO work without touching anything live.

Keep a human on every execution decision

New page creation, title tag and H1 changes, redirects, canonical tags, internal linking changes, and anything that touches a live URL get reviewed and approved by a person who actually knows the site before it ships. Not “spot-checked after the fact.”

Reviewed before it goes live, the same way you’d review a junior team member’s first few weeks of work, because that’s functionally what an AI agent with content permissions is.

Treat every AI-drafted ‘new page’ as a claim, not a fact, until you’ve verified it

The specific failure I hit, cloning instead of creating, is easy to catch if you check for it and easy to miss if you don’t. Before publishing anything an AI model produced in response to a keyword-targeting request, diff it against your existing content.

If the body copy matches an existing page by more than a sentence or two, it’s not a new page. It’s a duplicate wearing a new title tag. Shipping it will cost you rankings on both URLs instead of gaining you one.

Don’t confuse ‘the output looks done’ with ‘the output is correct’

This is the mental shift that matters most. An AI model has no internal signal that tells it, “I don’t actually have a good answer to this.”

It just produces the best-looking answer it can generate, which is sometimes a real answer and sometimes a page-shaped object that technically satisfies the prompt. Your job, or your team’s job, is to be the check that catches the difference before a search engine has to.

Dig deeper: 6 content audit workflows to build in Claude

Frequently asked questions

Can I let an AI agent publish SEO content automatically?

You can, technically. Whether you should depends entirely on whether you’re comfortable with content going live that nobody reviewed for accuracy, uniqueness, or fit with your existing pages.

Mine and Caleb’s stories both involve exactly that scenario: An AI model given the ability to act without a review step in between.

Why does Claude duplicate pages instead of writing new ones?

Nobody outside Anthropic can say with certainty what’s happening inside the model on any given run, but the observable pattern is consistent. When asked to solve a content problem quickly, the model reaches for the fastest plausible output.

Modifying an existing page is faster and lower-risk, from the model’s perspective, than generating new, differentiated content, even though the result is worse for SEO.

Is this specific to Claude, or do other AI models do it too?

The duplication error in this article occurred specifically with Claude on two separate sites. But the broader pattern, confident wrong answers and execution shortcuts that look complete but aren’t, shows up across ChatGPT, AI coding agents, and other models in the examples cited above. This is a category problem with autonomous AI execution, not a single vendor’s bug.

Keep humans responsible for what goes live

Claude made my SEO research faster this year. It also tried to solve a keyword problem twice by cloning a page and calling it done, and one of those batches made it onto a live site before someone caught it.

If your team is running AI-assisted SEO right now, ask one question this week: Does every AI-touched page still have a human’s name on the approval before it goes live?

Most “use AI for SEO” advice starts with prompts for writing content faster.

But speed isn’t the biggest opportunity. AI can help with the SEO work that’s harder to scale:

• Testing what works.
• Finding gaps in your topical coverage.
• Connecting data.
• Building useful tools.
• Uncovering stories worth pitching.

The adoption data shows just how much room there is to do more.

Why use AI for SEO at all?

Future-thinking SEOs are already using AI, and the adoption data shows exactly where. In Semrush’s survey on how marketers use AI for SEO, the top uses are the commodity tasks:

• 60% use it for keyword research.
• 48% for brainstorming content ideas.
• 38% for content briefs.

The strategic work sits at the bottom of the list.

• Only 18% use AI to plan topic clusters.
• 15% to find internal linking opportunities.
• Just 11% for SERP or content gap analysis.

That distribution shows where the opportunity is. Most marketers have pointed AI at the work everyone else is already automating, which produces more content but no advantage.

Won’t I get penalized for using AI in my SEO?

No. Google has said plainly that using AI to produce content isn’t against its guidelines, as long as the content is helpful and made for people.

Its systems reward quality regardless of how the page was produced and demote content built to game rankings rather than help the reader.

The key is making every page and every action meaningful and valuable to the user. A page technically can be entirely AI-generated and still win. The art is training your systems and your workflow to create that value on every page.

Here’s where that pays off, with a prompt you can run in ChatGPT, Claude, or any AI assistant today.

1. Build a gated content system

Run content through AI as a series of gates (idea, keyword research, brief, draft, fact and quality check, humanizing pass) where nothing reaches publish until it clears each one.

The failure mode of AI content is the firehose: hundreds of pages, no gates, all landing and creating an average piece of work. Google holds a patent on measuring information gain, the new information a page adds beyond what’s already indexed, and its systems reward pages that add rather than repeat. 

At my agency, every campaign carries an element of information gain somewhere:

• In the content itself.
• In digital PR that puts proprietary data into the world.
• In a unique angle.
• In an interactive tool.

Gates are where you force that gain in before the page ships.

What to do

Break the workflow into discrete stages and put a check at each one. The gate that matters most sits before drafting: Does this page add something that the top 10 search results don’t already have?

If not, it goes back for proprietary data or unique perspectives.

Simple AI prompt

What it returns

A go or no-go on the brief, with the exact evidence the page needs before it’s worth writing. It stops you from generating content that’s plain average and clearly AI-generated.

How to implement it

Run each stage as its own step rather than one prompt that writes end to end. I run this as a gated content workflow where idea, research, brief, draft, and humanizing are separate checks, and the information-gain gate is the final check before it launches live. 

The limit: Automate the steps, keep the judgment human. The gate is only as good as the person reading its output, so you do need to go back, read, and polish.

Dig deeper: How to build an AI content workflow from the ground up

2. Run your SEO experiments on autopilot

Pointing an autonomous AI loop at real SEO work: one scheduled session a day that reads its own memory, picks a single justified action per site, ships it inside hard guardrails, and gets scored honestly on one metric. Mine runs for under $5 a session.

I’ve done SEO for over 10 years, and the constant struggle is finding, in black and white, what actually worked. Tracing the smallest change to the ROI it produced is practically impossible by hand.

An autonomous loop with one metric per site and an honest scoring rule turns that into an experiment you can finally read.

What to do

Give the loop three things:

• A steering document (objectives, the evidence it may use, and hard guardrails).
• A warm-start memory (a state file and an append-only run log, so it doesn’t start blind each day).
• Exactly one metric per site.

Then let it choose one action per day. Building a page is one option, as is fixing a schema gap or deciding the best move today is to write a recommendation and ship nothing.

Simple AI prompt

What it returns

One justified action a day, and a record. The rule is the whole point: A metric that moved without a provable, page-specific cause doesn’t count.

On one run, a target set improved from an average position of 48 to 39, but the shipped fix had touched pages the metric doesn’t measure, so it was logged inconclusive rather than booked as a win.

A loop that can catch itself lying is worth more than one that always reports success.

How to implement it

A scheduled cloud session fires once a day and writes everything back to memory, then pushes it, because the push-back is the compounding mechanism: without it, tomorrow starts blind. Keep scoring separate from building. 

The loop ships, I score the metrics myself, on a schedule, so nothing self-grades. 

The limit: The loop makes the actions, you own the metric and the rules. The scoring has to be a human judgment because, on its own, an LLM can get sidetracked with things that don’t matter.

Dig deeper: Technical SEO testing: How to build a stronger experiment

3. Diagnose and close your topical map

Using AI to read what Google currently classifies your site as, then map the coverage gaps across your whole sitemap and your competitors’, so you build against the classification instead of guessing.

“Build topical authority” gets misread as “publish more content.” The real job is getting classified by Google and AI engines as the source for the topics that make you money, then compounding coverage on that classification. Publishing before you know your classification builds on a bad foundation.

What to do

Feed AI your crawl, your ranked keywords, and a few competitors’ sitemaps. Have it read back the classification, name the gap between that and what you want to own, and produce the prune list and the topical map.

Simple AI prompt

What it returns

A one-line verdict on what your site is “about,” the gap, a list of pages to remove, and the competitor topics you’re missing.

Reading the whole inventory at once also surfaces the compounding technical faults a human never scrolls to.

How to implement it

We run this as a staged build (crawl, diagnose, map, schedule) and a standing sitemap intelligence task that reads the full inventory against competitors on a cadence. 

The limit: Pacing decides the outcome. AI makes the map in an afternoon, which is exactly why the discipline has to be human. A new section matures for about three months before you scale into it, and a full topical network takes seven or eight months to publish at an irregular cadence. 

Publish a new section too fast and Google treats the whole thing as mass-produced content, then buries it. We’ve watched that wipe out most of a site’s traffic. Being a big brand doesn’t save you.

Dig deeper: Why every SEO team now needs a social topical map

4. Triangulate GSC, GA4, and Google Trends in one pass

Using AI to merge your Search Console, Analytics, and Google Trends data into one synthesis, instead of reading three siloed tabs and missing the overlap.

The answer usually lives in the overlap. A query rising in Trends, stuck at Position 8 in Search Console, with low engagement in Analytics, is a packaging problem. It looks like nothing in any single tool. Reading these sources separately makes it easy to miss how the data connects.

What to do

Export all three for the same pages and date range, hand them to AI together, and ask for the story that only appears when they’re read as one.

Simple AI prompt

What it returns

A prioritized list of moves that no single tool would surface: the rising-demand page with a weak title, the high-engagement page nobody can find, the query dropping before anyone noticed. We run this as a cross-source sweep that also pulls Bing, Clarity, and where the brand surfaces in AI search. 

The limit: Joining the data together. These sources don’t share a key or a scale:

• Search Console is query-level.
• Analytics is page-level.
• Google Trends is a relative 0-to-100 index that only makes sense within a single pull, not a number you can add to others.

AI lines them up and surfaces the pattern, but the result is an interpretation, so you confirm the story in the raw data before you act on it.

5. Build the interactive tools

Using AI to build calculators, templates, and interactive tools in raw HTML, fast, without a developer. Interactive tools are great for user experience, and they target high-intent searches: “X calculator” and “X template” are actionable, buying-adjacent queries.

Yet almost nobody builds them because teams assume it needs developer work. AI removes that barrier, which turns one of the most useful formats into one of the most neglected.

What to do

Take a calculation or decision your audience actually makes, describe the inputs and the logic, and have AI generate a self-contained tool you can embed.

Simple AI prompt

What it returns

A working, embeddable calculator you can ship as a lead magnet. Feed it your own data or formula to make it unique and brand-specific, which also helps it get cited by AI engines rather than ignored. 

The limit: AI writes the tool, the proprietary logic or dataset inside it is what makes it worth using.

6. Mine your data for digital PR angles

Using AI to find the newsworthy angle in your niche and the data to prove it, so a digital PR campaign earns the high-tier coverage that moves the needle in SEO.

The hardest part of digital PR is the angle, and it hides in what your audience complains about. Coverage from high-authority publications is also what search engines pull into their answers, so one data-led campaign earns links, citations, and AI-referred sales at once.

Most teams still treat PR as a brand-perception play, when it’s now one of the strongest inputs to AI search visibility.

What to do

Use AI in the recon. Mine Google News, Google Trends, Reddit, and niche forums for the recurring complaint, then shape it into a data angle you can prove with a simple comparison.

Build the proof, then a tight media list, then a short pitch a journalist can paste straight into a story.

Simple AI prompt

What it returns

A newsworthy angle, the data to prove it, and the people to pitch. One example: for the luggage brand Kadi, recon surfaced that travelers felt tricked by hidden airline fees.

A clean comparison of base fare versus final fare across the top 10 Australian airlines produced the headline “Australian airlines charge up to 66% more in hidden fees,” picked up by Yahoo Finance, Australian Traveller, escape.com.au, and more.

How to implement it

Those tier-one placements got pulled into ChatGPT, Perplexity, and Google’s AI Mode, where they keep answering “which airlines have the most hidden fees” long after the headlines fade. Run the recon and the media-list build with AI. 

The limit: The journalist relationship and the judgment of which angle is genuinely newsworthy stay human, and anything AI pulls from your data has to be double-checked before it reaches a journalist. A wrong figure in a pitch becomes a wrong figure in a headline, with your name on it.

Dig deeper: 7 digital PR secrets behind strong SEO performance

7. Map one topic across search, social, and video

Using AI to turn the queries your social content already ranks for into a single content plan across search, social, and video, a social topical map.

Search Console’s platform properties now show the Google queries your YouTube, TikTok, and social content appear for. On our own channel, that surfaced more than 200,000 impressions in 11 days from videos never optimized for Google. 

AI Mode is multimodal, so one query can hand a brand a page, a video, and a social post on the same results page. Mapped across formats, you hold several slots. Publishing only text, you hold one.

What to do

Cluster the platform-property queries, then run a query fan-out: The broad topic becomes the blog pillar, a facet becomes the long-form video, and a specific sub-question becomes the short. One cluster, several assets, all pulling on the same demand.

Simple AI prompt

What it returns

A cluster-by-cluster plan with a matched blog, video, and short for each. Embed the video on the matching page so the two reinforce each other for the same query. 

The limit: AI packages the plan. The judgment of which cluster is worth the bet stays yours. Social is a new and emerging play, and this data is brand new, so expect clearer patterns to show across industries and niches over the coming year.

Dig deeper: Google Search Console now connects social content to search demand

The pattern across all seven AI uses for SEO

Look back at the seven, and the shape is identical. None of them is a “quick AI hack.”

Each one points AI at work that compounds: gating quality in before it ships, running honest experiments, diagnosing what Google thinks you are, triangulating your own data, building the tools nobody else bothers with, mining your data for the story journalists want, and mapping one topic across every surface.

The adoption data shows most teams aiming AI at the commodity tasks, keyword research, and first drafts, the work everyone else is already automating. That makes more content and no advantage, and the parts that decide the outcome, the judgment, the proprietary data, and the relationships, are the parts you keep human.

Point AI at the leverage, keep your hands on the judgment, and it’ll compound into 2027 and beyond.

Link building in 2027 will be less about how many links you can acquire and more about whether you’ve earned them through stories worth covering. 

Links still matter, but the value of an earned placement extends to brand mentions, publication audiences, social discovery, and the authority that comes from being associated with relevant stories.

That changes what winning looks like. The strongest digital PR programs build genuine journalist relationships, create useful and original content, and find the story inside a brand, product, or data set. The weakest chase volume — mass pitches, bought links, recycled ideas, and content created simply to keep publishing.

Here are 13 ways to win or fail at link building in 2027.

Fails: Common link building mistakes to avoid

1. Flooding inboxes with bad AI pitches

AI isn’t the problem here. Plenty of teams are using it to move faster, personalize more, and test more angles than they ever could by hand. The problem is what happens when that speed comes without a strategy behind it.

Pitch volume has exploded because generating one no longer requires much effort, and journalists are drowning in emails that all sound like they were written by the same robot. The answer to that noise isn’t more noise.

The distinction that actually matters here is strategy versus volume, and volume lost this game years ago. If your 2027 plan is “AI writes 500 pitches a week,” you’re not scaling outreach, you’re scaling deletion, and worse, you’re burning sender reputation with journalists that took years to build. 

The teams winning with AI right now aren’t skipping the thinking. They’re doing more of it, faster, and using that speed to replicate structures that already worked instead of mass-producing first drafts of nothing and calling it a pitch.

2. Buying spammy links

Buying links can look good on paper (i.e., high domain ratings and seeing a large volume of links). But none of that reflects the real audience, editorial standards, or topical relevance. 

The second wave of Google’s March 2026 spam update targeted exactly this pattern, impacting sites that acquired authority through link manipulation rather than earning it. The brands hit hardest almost always shared the same trait in their backlink profile: links that were bought, not earned.

This should no longer be part of anyone’s strategy in 2027. It was always a risk and rarely holds up over time, delivering no lasting value: no journalist relationships, no genuine audience, and no signal that AI systems weigh when determining visibility.

3. Chasing speed over quality

AI has made it possible to create content and send pitches faster than ever, and that speed feels like an advantage until you look at what it’s actually producing. 

A pitch drafted in 30 seconds reads like it was drafted in 30 seconds, and a piece of content rushed with no thought, plan, or strategy behind it has the same effect.

Being first isn’t the same as being right, and being fast isn’t the same as being useful. A slower, sharper pitch and a slower, sharper piece of content will beat something rushed out to hit a deadline or catch a trend before it passes, every time. Not every piece of content should be fast, nor should every pitch. Overall, more thorough pitches and content will do better 100% of the time.

It’s about quality over quantity. The teams that win in 2027 will be the ones willing to sit on something, plan, and understand what will work best in the media landscape, not the ones bragging about how fast they can turn it around without a set plan.

4. Cannibalizing your own great content

The pressure to keep publishing often leads teams to create content that has no real reason to exist: no news hook, no data behind it, and no journalist who would ever pick it up.

Your content should be helpful to journalists, readers, and customers. If it serves none of them, it shouldn’t exist.

Every piece of content should answer one question: What is this actually earning us — a mention, a link, a share, a follower, a lead? If the answer is nothing, it isn’t marketing. It’s noise that can cannibalize your own great content.

We’re seeing people in the same space make the same piece of content 10 times over simply to maintain volume. It’s better to have one well-thought-out piece that gets picked up year over year than 10 pieces with no value.

Dig deeper: Why the traditional link building model no longer works for AI search

5. Copying competitors’ content verbatim

Competitor backlink profiles are still a great place to start, since they show you exactly where the gaps are and which publications are already covering your space. 

The losing move is to stop there and replicate what already worked for someone else instead of using that gap to build something better or build a similar idea that’s much improved.

In a recent report, 66.6% of SEO professionals said finding unique opportunities beats replicating competitors’ profiles. Your data, your experts, and your press history are inputs nobody else can clone, and that’s exactly what should fill the gap you just found, not a copy of what they already did. 

This is a great place to generate ideas, identify journalists, or note topics of interest, rather than simply copying content, which could also lead to cannibalization.

6. Missing the story entirely

A press release announcing a new hire, a product launch, or a funding round isn’t a story. It is an announcement, and journalists can tell the difference in the first sentence. 

The same goes for CEO profile pitches that spend three paragraphs on a founder’s background and vision and never once get to why any of it matters to a reader right now.

Talking about how great your product is will never be the story either, no matter how well it’s written. A journalist doesn’t care that your product is great. They care whether it solves a real problem, reveals something surprising, or connects to something already happening in the news. 

The goal was never to describe the product, the brand, or the executive. It was always about finding the actual story sitting inside them and leading with that instead.

If a pitch can’t answer why a journalist would care, it doesn’t matter how polished the press release is, how impressive the CEO sounds, or how many features get listed. Without the story, none of it goes anywhere.

7. Not using AI correctly

AI is scary right now, and much of that fear stems from how it’s being discussed and how quickly it’s been introduced into the workplace. It makes sense why companies are excited about the efficiencies AI can bring, especially when teams are being asked to move faster, create more, analyze more, and keep up with an industry that’s constantly changing.

When used correctly, AI helps companies create stronger content, derive data insights, and iterate faster, building consistency into processes, sharpening decision-making, and easing the team’s cognitive load. 

There’s a reason people feel nervous about AI. When the conversation is centered solely on automation, speed, and doing more with less, it can start to sound as though the human side of the work is being pushed aside. 

The best use of AI isn’t replacing people. It is reducing some of the cognitive load that comes with the work and giving teams more time to focus on the bigger strategic priorities that actually need them. Going AI-first should mean putting your people first, not the other way around.

Dig deeper: Link building works better when you think beyond links

Wins: Effective link building strategies

8. Run one earned media motion

Most companies still split this work across silos: an SEO team building links, a PR team landing traditional press, podcasts, and appearances, and a brand team running socials with executive storytelling. 

Heading into 2027, that wall is a structural failure because every earned placement can contribute a link signal, a brand mention, and a potential LLM citation source.

The teams that merge these into one digital PR motion, with shared targets, shared story assets, and shared credit, get triple value from every earned media mention, while siloed teams run separate goals without cashing in on the full value repackaging marketing content can bring.

9. Treat journalist relationships as your most defensible asset

Great PR is a bridge to your most defensible asset: journalist relationships.

So why are we still sending transactional, en masse, uninteresting stories? Why are we treating journalists like there isn’t a human behind the computer screen?

Over the last seven years, we’ve sent over 10,000 pitches and landed hundreds of pieces of coverage. It hasn’t been about being transactional. Every pitch has a story and value, and it’s helpful for a future story. 

Being a bridge starts with doing the research: noting what journalists cover, what they’ve written recently, and what they genuinely care about so the relationship can go both ways.

Plainly, you aren’t owed coverage because you sent a nice email, and a journalist doesn’t owe you a read because you did your homework. 

You’re building something mutually beneficial over time, where you keep tabs on socials, meet for coffee chats, and consistently read a journalist’s coverage. Even more, you can reference podcasts or other appearances to learn as best you can in order to send only the most relevant pitches.

Additionally, once that relationship is established, don’t simply send pitches just to get coverage. This can ruin the relationship. Do your homework, create helpful content, and treat journalists as your most important asset.

10. Trying several tactics and experimenting with your resources

Data studies, product launches, expert commentary, and social media aren’t competing tactics. They’re a portfolio, and if PR teams are experimenting with these now or with other content types, you could be left out of great press opportunities.

Survey campaigns are surging, but the space is getting saturated as brands run the same types of surveys instead of chasing something unique or novel. A good survey ties to the news cycle and has broad appeal, depending on the topic, sample size, and timing of the pitch.

However, without those key pillars, it’s just another brand’s noise competing with everyone else’s noise in the same inbox. Try other things, and don’t be afraid to experiment or mix and match strategies. 

For example, product stories empower the brand itself, expert commentary keeps your spokespeople quotable when news breaks in your territory, and social media expands your reach.

Muck Rack’s State of Journalism 2026 report found that access to credible sources is one of the things journalists value most in a pitch, making expert commentary the most underrated LLM play of the three. A named expert who is repeatedly quoted on a topic can become an entity AI systems associate with that topic. Additionally, more than half of Americans get their news from social media.

Dig deeper: From links to brand signals: The new SEO authority model

11. Redefining your metrics of success

Volume will matter less in 2027. SEO is changing, and we’re seeing brands penalized for spammy links, overreliance on volume, and the use of fake experts to inflate volume. What will matter in 2027 is how your brand is mentioned, in what context, and how many times. What’s the point of 50 links with nothing contextual for your brand attached to them?

Look at where brands are winning right now: TikTok, LinkedIn, and genuinely contextual coverage. Brand mentions still matter, but context matters more than count. The goal is to earn mentions that can help your brand get surfaced by the AI systems shaping discovery.

Your 2027 scorecard should reflect that, not just how many times your name showed up somewhere, but where, and whether it made any sense for it to be there.

12. Duplicating what already worked

Your best pitch from last year is your best template for next year, and it’s probably sitting in your sent folder collecting virtual dust. 

Take the subject line pattern, the opening hook, the stat sequencing, and the reader-first call to action from your biggest win, and use AI to replicate that structure for every new campaign instead of prompting from a blank page. The structure is yours, built from your wins, your voice, and your relationships, so it scales without going generic.

Our team has seen great success with this format, and it works for content as well. Our best-performing pieces were built on the exact structures that already worked at that level, rather than generated from a blank prompt.

That’s exactly the gap most PR teams are still missing, and incorporating what I call the digital PR duplication method could eliminate the need to consistently reinvent the wheel.

Sometimes, the best way to start is by noting what has worked in the past. You’re compounding, and the whole point is to rinse, reuse, and repeat.

13. Thinking like a journalist

Before creating any content, pitch, or campaign idea, put your journalist cap on. That means asking the same questions a journalist would ask before running a story:

• Why does this matter right now? 
• Who does it actually affect? 
• What angle has nobody else already covered?

That mindset should shape the entire story: the data you decide to pull, the product update you build a campaign around, the expert commentary you put together, and the social post that supports the placement once it lands.

A lot of PR teams jump straight to “how do we say this” without ever asking “would anyone actually want to read this,” and that’s where most of what they make loses before it ever goes out the door.

It leads with the story, not the brand. It anticipates the follow-up question before it gets asked. It shows up with the stat already pulled, the quote already tight, and the context already there because the person creating it already did the work a journalist or a reader would otherwise have to do themselves. 

Every campaign idea, every pitch, and every piece of content should survive one question before it goes anywhere: Would a journalist or a reader care about this?

2027 is when digital PR stops hiding behind links

For years, digital PR has hidden behind the link because it’s easy to report and has often been treated as the primary measure of success. In 2027, the focus needs to be broader: real relationships, topical authority, and stories editors actually want — the things authority is made of.

That value is earned in two ways:

• Quality over quantity: One well-thought-out campaign, pitch, or piece of content can outperform five mediocre ones, whether it’s earned media, social content, a speaking opportunity, or a podcast appearance. That’s especially true as AI slop, content farms, and shrinking newsrooms make attention harder to earn.
• Being a bridge, not a burden: Build relationships with the journalists, clients, and audiences who give that authority meaning. Give them something useful, relevant, and worth their attention.

That’s what winning looks like in 2027. It’s not about counting the most links. It’s about building something worth being known for.

Google’s site reputation abuse manual action will no longer impact the search results in the European Economic Area (EEA) starting August 30, 2026. This is in response to the European Commission’s investigation over the site reputation abuse policy and its requiring Google to adjust its systems for that policy not to be enforced within the EEA.

What is changing. On August 30th, manual actions applied under Google’s site reputation policy will have different effects for those searching in the EEA than for those outside it. This is how it will differ, says Google:

• For users outside the EEA, a manual action regarding Google’s site reputation policy will directly affect search results for the portion of the site affected. As before, the rest of the site won’t be affected.
• For users inside the EEA, the impact of the manual action will not apply. The affected section of the site may be separated in our systems so that, over time, it ranks independently from the rest of the site.

It seems like Google’s site reputation abuse manual actions won’t specifically apply within the EEA but its systems for treating the portion of the site that hosts third-party content as a “parasite SEO” strategy will be treated as its own site and not as part of the existing site. The concept of Google treating sections of a site as its own site is not new to Google’s site reputation abuse spam prevention, and that still seems to be happening in the EEA regions. But manual actions over site reputation abuse will not apply in the EEA regions.

What Google said. A Google spokesperson told us:

“Our European users are no less frustrated by parasite SEO and other deceptive, pay-to-play tactics that degrade search results, and we stand by our Site Reputation Policy. While we remain concerned about efforts to weaken our spam policies, we’ve agreed to make changes to our enforcement approach for users in Europe and clarify our policies to address the European Commission’s concerns.”

More details. In March 2024, Google released the March 2024 core update, along with several spam policies including the new site reputation abuse policy. Google’s anti-spam policies are essential to how Google fights deceptive pay-for-play tactics, aka “parasite SEO,” that can degrade Google’s search results.

The adjustment Google announced specifically applies to how Google handles manual actions under the Site Reputation Policy (SRAP) for search results displayed to users inside the European Economic Area (EEA). Google is not pausing or weakening its core spam enforcement in Europe.

Manual action notifications will still be sent to sites based in the EEA region over site reputation abuse but those penalties will not impact the site’s placement in Google’s search results for searchers within the EEA region. It will affect what searchers see outside the EEA.

Google wrote, “Site owners will continue to be notified within Search Console when a manual action is applied. If you believe an action was taken in error, you can submit a reconsideration request. Eligible sites will, following the reconsideration request, also have the opportunity to bring disputes to mediation. Because many pages are viewed by people globally, it’s possible that any given page might have a manual action taken on it, but that change will only affect search results shown to users outside the EEA.”

Why we care.

If your website or client’s site has a site reputation abuse manual action and it gets Google search traffic from searchers in the EEA region, than you might notice even more traffic after this rolls out.

It is still recommended that if you want to rank well outside of the EEA, you do not violate the site reputation abuse policy.

I am a bit surprised that Google actually is changing how it handles a spam policy based on the region of the searcher but I guess Google decided it has no other choice.