Another year in search has come and gone, and Google called it year three of a 10-year platform shift. In 2025, that shift became impossible to ignore. AI moved from experiments and previews into the core of how search actually works.
Below are the biggest SEO news stories of 2025 on Search Engine Land.
Note: This article doesn’t include any stories related to Google algorithm updates. Barry Schwartz wrote a separate recap on that, which will also publish today.
10. Perplexity ranking factors and systems
Independent researcher Metehan Yesilyurt analyzed browser-level interactions to reveal how Perplexity scores, reranks, and sometimes drops content. He uncovered a three-layer machine-learning reranker for entity searches, manual authority whitelists, and dozens of engagement and relevance signals.
Yesilyurt’s research also found boosts for authoritative domains, strong early performance, and topics centered on tech and AI. Rankings further reflected time decay, interconnected content clusters, and synchronized YouTube trends that increased visibility across platforms.
Your customers search everywhere. Make sure your brand shows up.
The SEO toolkit you know, plus the AI visibility data you need.
Start Free Trial
Get started with
9. Google Search Console Query groups
Google added Query groups to the Search Console Insights report. The feature uses AI to cluster similar search queries into clear audience topics and does not affect rankings. It rolled out gradually to high-volume sites and replaced long query lists with topic-level groupings that make performance shifts easier to spot.
HubSpot’s organic traffic appeared to fall from 13.5 million to 8.6 million in a month, with most of the losses coming from its blog. The drop followed several Google updates, and SEOs publicly pointed to thin, off-topic, traffic-at-all-costs content that drifted beyond HubSpot’s core expertise.
The SEO identity crisis continued as Google dismissed new acronyms like GEO (generative engine optimization) and AEO (answer engine optimization), arguing that good SEO is good GEO, and that the same fundamentals drive AI Overview rankings.
That stance collided with Google’s own admission that search traffic decline is inevitable as AI answers replace clicks, even while traditional search still dominates discovery at a massive scale.
Yet, search behavior is fracturing: users turn to AI for quick answers and to Google for deeper research, pushing brands to optimize for visibility, not just traffic.
Google rapidly expanded AI Mode from an opt-in experiment into a widely available, and possibly soon default, search experience. It added deeper research, agentic actions, personalization, and Gemini 2.5, signaling longer and more complex search behavior.
At the same time, AI Mode exposed major transparency gaps. It initially broke referral tracking and still blends performance data into standard Search Console reports, raising new concerns about visibility, attribution, and what SEO becomes as AI takes on a larger role in search.
Cloudflare CEO Matthew Prince said AI was breaking the web’s search-driven business model. He said Google scraped far more content while sending back much less traffic because of zero-click results. He added that AI companies deepen the imbalance by consuming huge amounts of content with little return to creators, putting original publishing at risk unless the economic model changes.
Statcounter data showed Google’s global search share fell below 90% in October, November, and December 2024, the first time its search share remained under 90% since early 2015. The decline was driven mainly by Asia, alongside a December U.S. dip to 87.39%. Bing, Yandex, and Yahoo captured much of the lost share.
Google tightened its stance on AI-generated content by telling quality raters to give the Lowest ratings to pages where most main content is auto- or AI-generated with little originality or added value. It also expanded its spam definitions to target scaled, low-effort AI use.
At the same time, Google tested AI-generated and AI-summarized search snippets, pointing to a future where AI both judges content more harshly and increasingly controls how that content appears in search.
Analyses from Seer, Ahrefs, Amsive, and BrightEdge all showed the same pattern. Google Search produced more impressions and more AI Overview visibility, but sent fewer clicks. The drop was sharpest on non-branded, informational queries, where AI Overviews pushed classic results down, and CTR fell hard.
The studies also found a winner-take-some dynamic. Brands cited in AI Overviews saw higher paid and organic CTR, while those left out lost ground, showing that AI visibility increasingly drives results.
Google’s removal of the long-standing &num=100 search parameter disrupted SEO data across the industry. It broke rank-tracking tools and coincided with sharp drops in Google Search Console impressions and query counts.
Early analysis showed most sites lost reported visibility, especially beyond Page 1. The change suggested years of inflated metrics from scrapers and a new, possibly more accurate, view of organic performance.
https://i0.wp.com/dubadosolutions.com/wp-content/uploads/2021/12/web-design-creative-services.jpg?fit=1500%2C600&ssl=16001500http://dubadosolutions.com/wp-content/uploads/2017/05/dubado-logo-1.png2025-12-29 13:00:002025-12-29 13:00:00Top 10 SEO news stories of 2025
OpenAI is laying the groundwork for an advertising business, signaling a potential shift in how ChatGPT and other products could be monetized beyond subscriptions and enterprise deals.
What’s happening. According to reporting from The Information, OpenAI has begun exploring ad formats and partnerships, with early discussions pointing toward ads that could appear within or alongside AI-generated responses. The effort is still in its early stages, but internal conversations suggest ads are becoming a more serious part of OpenAI’s long-term revenue strategy.
Why we care. OpenAI is exploring ads inside AI-generated responses, creating a new, highly contextual channel for reaching users at the moment they seek information. This could put OpenAI in direct competition with Google and Meta, but also raises questions about trust and user engagement. Early adoption could offer a first-mover advantage, while formats and metrics may differ from traditional digital ads. Overall, it’s a potentially transformative new frontier for advertising.
Your customers search everywhere. Make sure your brand shows up.
The SEO toolkit you know, plus the AI visibility data you need.
Start Free Trial
Get started with
Between the lines. OpenAI appears cautious, aiming to avoid disrupting user experience or undermining confidence in its models. Any ad product is likely to be tightly controlled, at least initially, and positioned as helpful or contextually relevant rather than overtly promotional.
The bigger picture. With soaring infrastructure costs and growing pressure to scale revenue, ads could become a key lever for OpenAI — especially as generative AI reshapes how people search for information and discover products.
What to watch. When ads move from internal planning to public testing, how clearly they’re labeled, and whether users accept advertising embedded in AI responses.
Bottom line. OpenAI isn’t rushing ads to market, but the foundations are being laid — and their eventual arrival could reshape both AI products and the digital advertising landscape.
https://i0.wp.com/dubadosolutions.com/wp-content/uploads/2021/12/web-design-creative-services.jpg?fit=1500%2C600&ssl=16001500http://dubadosolutions.com/wp-content/uploads/2017/05/dubado-logo-1.png2025-12-24 16:18:022025-12-24 16:18:02OpenAI discusses an ad-driven strategy centered on ChatGPT scale and media partnerships
Google reduced the minimum audience size requirement to just 100 active users across all networks and audience types, making remarketing and customer list targeting far more accessible—especially for smaller advertisers.
What’s new. Audience segments with as few as 100 users can now be used across Search, Display, and YouTube, including both remarketing lists and customer lists. The same 100-user threshold now applies for segments to appear in Audience Insights, down from 1,000.
Why we care. Smaller accounts and niche advertisers can now activate audience strategies that were previously out of reach due to size constraints. This change removes a long-standing barrier to personalization, remarketing, and first-party data activation within Google Ads.
What to watch. How advertisers use smaller, more precise segments—and whether performance or privacy safeguards evolve alongside the expanded access.
First seen. This update was first spotted by Web Marketing Consultant, Dario Zannoni, who shared it on LinkedIn.
SEO didn’t stand still in 2025. It didn’t reinvent itself either. It clarified what actually matters. If you followed The SEO Update by Yoast monthly webinars this year, you’ll recognize the pattern. Throughout 2025, our Principal SEOs, Carolyn Shelby and Alex Moss, cut through the noise to explain not just what was changing but why it mattered as AI-powered search reshaped visibility, trust, and performance. If you missed some sessions or want the full picture in one place, this wrap-up is for you. We’re looking back at how SEO evolved over the year, what those changes mean in practice, and what they signal going forward.
Key takeaways
In 2025, SEO shifted its focus from rankings to visibility management, as AI-driven search reshaped priorities
Key developments included the rise of AI Overviews, a shift from clicks to citations, and increased importance of clarity and trust
Brands needed to prioritize structured, credible content that AI systems could easily interpret to remain visible
By December, SEO transformed to retrieval-focused strategies, where success rested on clarity, relevance, and E-E-A-T signals
Overall, 2025 clarified that the fundamentals still matter but emphasized the need for precision in content for AI-driven systems
AI-powered, personalized search accelerated. Zero-click results increased. Brand signals, E-E-A-T, performance, and schema shifted from optimizations to requirements.
SEO expanded from ranking pages to representing trusted brands that machines can understand.
February
Massive AI infrastructure investments. AI Overviews pushed organic results down. Traffic dropped while brand influence and revenue held steady.
SEO outcomes can no longer be measured by traffic alone. Authority and influence matter more than raw clicks.
March
AI Overviews expanded as clicks declined. Brand mentions appeared to play a larger role in AI-driven citation and selection behavior than links alone. Search behavior grew despite fewer referrals.
Visibility fractured across systems. Trust and brand recognition became the differentiators for inclusion.
April
Schema and structure proved essential for AI interpretation. Multimodal and personalized search expanded. Zero-click behavior increased further.
SEO shifted from optimization to interpretation. Clarity and structure determine reuse.
May
Discovery spread beyond Google. AI Overviews reached mass adoption. Citations replaced visits as success signals.
SEO outgrew the SERP. Presence across platforms and AI systems became critical.
June – July
AI Mode became core to search. Ads entered AI answers. Indexing alone no longer offers guaranteed visibility. Reporting lagged behind reality.
Traditional SEO remained necessary but insufficient. Resilience and adaptability became essential.
August
Visibility without value became a real risk. SEO had to tie exposure to outcomes beyond the number of sessions.
Visibility without value became a real risk. SEO had to tie exposure to outcomes beyond sessions.
September
AI Mode neared default status. Legal, licensing, and attribution pressures intensified. Persona-based strategies gained relevance.
Control over visibility is no longer guaranteed. Trust and credibility are the only durable advantages.
October
Search Console data reset expectations. AI citations outweighed rankings. AI search became the destination.
SEO success depends on presence inside AI systems, not just SERP positions.
November
AI Mode became core to search. Ads entered AI answers. Indexing alone is no longer a guarantee of visibility. Reporting lagged behind reality.
Clarity and structure beat scale. Authority decides inclusion.
December
SEO fully shifted to retrieval-based logic. AI systems extracted answers, not pages. E-E-A-T acted as a gatekeeper.
SEO evolved into visibility management for AI-driven search. Precision replaced volume.
January: SEO enters the age of representation
January set the tone for the year. Not through a single disruptive update, but through a clear signal that SEO was moving away from pure rankings toward something broader. The search was becoming more personalized, AI-driven, and selective about which sources it chose to surface. Visibility was no longer guaranteed just because you ranked well.
From the start of the year, it was clear that SEO in 2025 would reward brands that were trusted, technically sound, and easy for machines to understand.
What changed in January
Here are a few clear trends that began to shape how SEO worked in practice:
AI-powered search became more personalized: Search results reflected context more clearly, taking into account location, intent, and behavior. The same query no longer produced the same result for every user
Zero-click searches accelerated: More answers appeared directly in search results, reducing the need to click through, especially for informational and local queries
Brand signals and reviews gained weight: Search leaned more heavily on real-world trust indicators like brand mentions, reviews, and overall reputation
E-E-A-T became harder to ignore: Clear expertise, ownership, and credibility increasingly acted as filters, not just quality guidelines
The role of schema started to shift: Structured data mattered less for visual enhancements and more for helping machines understand content and entities
What to take away from January
January wasn’t about tactics. It was about direction.
SEO started rewarding clarity over cleverness. Brands over pages. Trust over volume. Performance over polish. If search engines were going to summarize, compare, and answer on your behalf, you needed to make it easy for them to understand who you are, what you offer, and why you are credible.
That theme did not fade as the year went on. It became the foundation for everything that followed.
February: scale, money, and AI made the shift unavoidable
If January showed where search was heading, February showed how serious the industry was about getting there. This was the month where AI stopped feeling like a layer on top of search and started looking like the foundation underneath it.
Massive investments, changing SERP layouts, and shifting performance metrics all pointed to the same conclusion. Search was being rebuilt for an AI-first world.
What changed in February
As the month unfolded, the signs became increasingly difficult to ignore.
AI Overviews pushed organic results further down: AI Overviews appeared in a large share of problem-solving queries, favoring authoritative sources and summaries over traditional organic listings
Traffic declined while brand value increased: High-profile examples showed sessions dropping even as revenue grew. Visibility, influence, and brand trust started to matter more than raw sessions
AI referrals began to rise: Referral traffic from AI tools increased, while Google’s overall market share showed early signs of pressure. Discovery started spreading across systems, not just search engines
What to take away from February
February made January’s direction feel permanent.
When AI systems operate at this scale, they change how visibility works. Rankings still mattered, but they no longer told the full story. Authority, brand recognition, and trust increasingly influenced whether content was surfaced, summarized, or ignored.
The takeaway was clear. SEO could no longer be measured only by traffic. It had to be understood in terms of influence, representation, and relevance across an expanding search ecosystem.
March: visibility fractured, trust became the differentiator
By March, the effects of AI-driven search were no longer theoretical. The conversation shifted from how search was changing to who was being affected by it, and why.
This was the month where declining clicks, citation gaps, and publisher pushback made one thing clear. Search visibility was fragmenting across systems, and trust became the deciding factor in who stayed visible.
What changed in March
The developments in March added pressure to trends that had already been forming earlier in the year.
AI Overviews expanded while clicks declined: Studies showed that AI Overviews appeared more frequently, while click-through rates continued to decline. Visibility increasingly stopped at the SERP
Brand mentions mattered more than links alone: Citation patterns across AI platforms varied, but one signal stayed consistent. Brands mentioned frequently and clearly were more likely to surface
Search behavior continued to grow despite fewer clicks: Overall search volume increased year over year, showing that users weren’t searching less; they were just clicking less
AI search struggled with attribution and citations: Many AI-powered results failed to cite sources consistently, reinforcing the need for strong brand recognition rather than reliance on direct referrals
Search experiences became more fragmented: New entry points like Circle to Search and premium AI modes introduced additional layers to discovery, especially among younger users
Structured signals evolved for AI retrieval: Updates to robots meta tags, structured data for return policies, and “sufficient context” signals showed search engines refining how content is selected and grounded
March exposed the tension at the heart of modern SEO.
Search demand was growing, but traditional traffic was shrinking. AI systems were answering more questions, but often without clear attribution. In that environment, being a recognizable, trusted brand mattered more than being the best-optimized page.
The implication was simple. SEO was no longer just about earning clicks. It was about earning inclusion, recognition, and trust across systems that don’t always send users back.
April: machines started deciding how content is interpreted
By April, the focus shifted again. The question was no longer whether AI would shape search, but how machines decide what content means and when to surface it.
After March exposed visibility gaps and attribution issues, April zoomed in on interpretation. How AI systems read, classify, and extract information became central to SEO outcomes.
What changed in April
April brought clarity to how modern search systems process content.
Schema has proven its value beyond rankings: Microsoft has confirmed that schema markup helps large language models understand content. Bing Copilot used structured data to generate clearer, more reliable answers, reinforcing the schema’s role in interpretation rather than visual enhancement
AI-driven search became multimodal: Image-based queries expanded through Google Lens and Gemini, allowing users to search using photos and visuals instead of text alone
AI Overviews expanded during core updates: A noticeable surge in AI Overviews appeared during Google’s March core update, especially in travel, entertainment, and local discovery queries
Clicks declined as summaries improved: AI-generated content summaries reduced the need to click through, accelerating zero-click behavior across informational and decision-based searches
Content structure mattered more than special optimizations: Clear headings that boost readability, lists, and semantic cues helped AI systems extract meaning. There were no shortcuts. Standard SEO best practices carried the weight
What to take away from April
April shifted SEO from optimization to interpretation.
Search engines and AI systems didn’t just look for relevance. They looked for clarity. Content that was well-structured, semantically clear, and grounded in real entities was easier to understand, summarize, and reuse.
The lesson was subtle but important. You didn’t need new tricks for AI search. You needed content that was easier for machines to read and harder to misinterpret.
By May, it was no longer sufficient to discuss how search engines interpret content. The bigger question became where discovery was actually happening.
SEO started expanding beyond Google. Visibility fractured across platforms, AI tools, and ecosystems, forcing brands to think about presence rather than placement.
What changed in May
The month highlighted how search and discovery continued to decentralize.
Search behavior expanded beyond traditional search engines: Around 39% of consumers now use Pinterest as a search engine, with Gen Z leading adoption. Discovery increasingly happened inside platforms, not just through search bars
AI Overviews reached mass adoption: AI Overviews reportedly reached around 1.5 billion users per month and appeared in roughly 13% of searches, with informational queries driving most of that growth
Clicks continued to give way to citations: As AI summaries became more common, being referenced or cited mattered more than driving a visit, especially for top-of-funnel queries
AI-powered search diversified across tools: Chat-based search experiences added shopping, comparison, and personalization features, further shifting discovery away from classic result pages
Economic pressure on content ecosystems increased: Industry voices warned that widespread zero-click answers were starting to weaken the incentives for content creation across the web
May reframed SEO as a visibility problem, not a traffic problem.
When discovery happens across platforms, summaries, and AI systems, success depends on how clearly your content communicates meaning, credibility, and relevance. Rankings still mattered, but they were no longer the primary measure of success.
The message was clear. SEO had outgrown the SERP. Brands that focused on authenticity, semantic clarity, and structured information were better positioned to stay visible wherever search happened next.
By early summer, SEO entered a more uncomfortable phase. Visibility still mattered, but control over how and where content appeared became increasingly limited.
June and July were about adjustment. Search moved closer to AI assistants, ads blended into answers, and traditional SEO signals no longer guaranteed exposure across all search surfaces.
What changed in June and July
This period introduced some of the clearest operational shifts of the year.
AI Mode became a first-class search experience: AI Mode was rolled out more broadly, including incognito use, and began to merge into core search experiences. Search was no longer just results. It was conversation, summaries, and follow-ups
Ads entered AI-generated answers: Google introduced ads inside AI Overviews and began testing them in conversational AI Mode. Visibility now competes not only with other pages, but with monetized responses
Measurement lagged behind reality: Search Console confirmed AI Mode data would be included in performance reports, but without separate filters or APIs. Visibility changed more rapidly than reporting tools could keep pace.
Citations followed platform-specific preferences: Different AI systems favored different sources. Some leaned heavily on encyclopedic content, others on community-driven platforms, reinforcing that one SEO strategy would not fit every system
Most AI-linked pages still ranked well organically: Around 97% of URLs referenced in AI Mode ranked in the top 10 organic results, showing that strong traditional SEO remained a prerequisite, even if it was no longer sufficient
Content had to resist summarization: Leaks and tests showed that some AI tools rarely surfaced links unless live search was triggered. Generic, easily summarized modern content became easier to replace
Infrastructure became an SEO concern again: AI agents increased crawl and request volume, pushing performance, caching, and server readiness back into focus
Search moved beyond text: Voice-based interactions, audio summaries, image-driven queries, and AI-first browsers expanded how users searched and consumed information
What to take away from June and July
This period forced a mindset shift.
SEO could no longer assume that ranking, indexing, or even traffic guaranteed visibility. AI systems decided when to summarize, when to cite, and when to bypass pages entirely. Ads, assistants, and alternative interfaces now often sit between users and websites more frequently than before.
The conclusion was pragmatic. Strong fundamentals still mattered, but they weren’t the finish line. SEO now requires resilience: content that carries authority, resists simplification, loads fast, and stays relevant even when clicks don’t follow.
By the end of July, one thing was clear. SEO wasn’t disappearing. It was operating under new constraints, and the rest of the year would test how well teams adapted to them.
August: the gap between visibility and value widened
By August, SEO teams were staring at a growing disconnect. Visibility was increasing, but traditional outcomes were harder to trace back to it.
This was the month when the mechanics of AI-driven search became more transparent and more uncomfortable.
What changed in August
August surfaced the operational realities behind AI-powered discovery.
Impressions rose while clicks continued to decline: AI Overviews dominated the results, driving exposure without generating traffic. In some cases, conversions still improved, but attribution became harder to prove
The “great decoupling” became measurable: Visibility and performance stopped moving in sync. SEO teams saw growth in impressions even as sessions declined
Zero-click searches accelerated further: No-click behavior climbed toward 69%, reinforcing that many user journeys now ended inside search interfaces
AI traffic stayed small but influential: AI-driven referrals still accounted for under 1% of traffic for most sites, yet they shaped expectations around answers, speed, and convenience
Retrieval logic shifted toward context and intent: New retrieval approaches prioritized meaning, relationships, and query context over keyword matching
It reinforced the reality that SEO could no longer rely on traffic as the primary proof of value. Visibility still mattered, but only when paired with outcomes that could survive reduced clicks and blurred attribution.
The lesson was strategic. SEO needed to connect visibility to conversion, brand lift, or long-term trust, not just sessions. Otherwise, its impact would be increasingly hard to defend.
September: control, attribution, and trust were renegotiated
September pushed the conversation further. It wasn’t just about declining clicks anymore. It was about who controlled discovery, attribution, and access to content.
This was the month where legal, technical, and strategic pressures collided.
What changed in September
September reframed SEO around governance and credibility.
AI Mode moved closer to becoming the default: Search experiences shifted toward AI-driven answers with conversational follow-ups and multimodal inputs
The decline of the open web was acknowledged publicly: Court filings and public statements confirmed what many publishers were already feeling. Traditional web traffic was under structural pressure
Legal scrutiny intensified: High-profile settlements and lawsuits highlighted growing challenges around training data, summaries, and lost revenue
Licensing entered the SEO conversation: New machine-readable licensing approaches emerged as early attempts to restore control and consent
Snippet visibility became a gateway signal: AI tools relied heavily on search snippets for real-time answers, making concise, extractable content more critical
Persona-based strategies gained traction: SEO began shifting from keyword targeting to persona-driven content aligned with how AI systems infer intent
Trust eroded around generic, formulaic, AI writing styles: Formulaic, overly polished AI content raised credibility concerns, reinforcing the need for editorial judgment
Measurement tools lost stability again: Changes to search parameters disrupted rank tracking, reminding teams that SEO reporting would remain volatile
What to take away from September
September forced SEO to grow up again.
Control over visibility, attribution, and content use was no longer guaranteed. Trust, clarity, and credibility became the only durable advantages in an ecosystem shaped by AI intermediaries.
The takeaway was sobering but useful. SEO could still drive value, but only when it is aligned with real user needs, strong brand signals, and content that earned its place in AI-driven answers.
October marked a turning point in how SEO performance needed to be interpreted. The data didn’t just shift. It reset expectations entirely.
This was the month when SEO teams had to accept that AI-powered search was no longer a layer on top of results. It was becoming the place where searches ended.
What changed in October
October brought clarity, even if the numbers looked uncomfortable.
AI Mode reshaped user behavior: Around a third of searches now involve AI agents, with most sessions staying inside AI panels. Clicks became the exception, not the default
AI citations increasingly rivalled rankings: Visibility increasingly depended on whether content was selected, summarized, or cited by AI systems, not where it ranked
Search engines optimized for ideas, not pages: Guidance from search platforms reinforced that AI systems extract concepts and answers, not entire URLs
Metadata lost some direct control: Tests of AI-generated meta descriptions suggested that manual optimization would carry less influence over how content appears
Commerce and search continued to merge: AI-driven shopping experiences expanded, signaling that transactional intent would increasingly be handled inside AI interfaces
What to take away from October
October reframed SEO as presence within AI systems.
Traffic still mattered, but it was no longer the primary outcome. The real question became whether your content appeared at all inside AI-driven answers. Clarity, structure, and extractability replaced traditional ranking gains as the most reliable levers.
From this point on, SEO had to treat AI search as a destination, not just a gateway.
November: structure and credibility decided inclusion
If October reset expectations, November showed what actually worked.
This month narrowed the gap between theory and practice. It became clearer why some content consistently surfaced in AI results, while other content disappeared.
What changed in November
November focused on how AI systems select and trust sources.
Structured content outperformed clever content: Clear headings, predictable formats, and direct answers made it easier for AI systems to extract and reuse information
Schema supported understanding, not visibility alone: Structured data remained valuable, but only when paired with clean, readable on-page content
AI-driven shopping and comparisons accelerated: Product data quality, consistency, and accessibility directly influenced whether brands appeared in AI-assisted decision flows
Citation pools stayed selective: AI systems relied on a relatively small set of trusted sources, reinforcing the importance of brand recognition and authority
Search tooling evolved toward themes, not keywords: Grouped queries and topic-based insights replaced one-keyword performance views
What to take away from November
November made one thing clear. SEO wasn’t about producing more content or optimizing harder. It was about making content easier to understand and harder to ignore.
Clarity beats creativity. Structure beat scale. Authority determined whether content was reused at all.
This month quietly reinforced the fundamentals that would define SEO going forward.
Instead of introducing new disruptions, it clarified what 2025 had been building toward all along. SEO was no longer primarily about ranking pages. It was about enabling retrieval.
What changed in December
The year-end review highlighted the new reality of SEO.
Search systems retrieved answers, not pages: AI-driven search experiences pulled snippets, definitions, and summaries instead of directing users to full articles
Literal language still mattered: Despite advances in understanding, AI systems relied heavily on exact phrasing. Terminology choices directly affected retrieval
Content structure became mandatory: Front-loaded answers, short paragraphs, lists, and clear sections made content usable for AI systems
Relevance replaced ranking as the core signal: Being the clearest and most contextually relevant answer mattered more than traditional ranking factors
E-E-A-T acted as a gatekeeper: Recognized expertise, authorship, and trust signals determined whether content was eligible for reuse
Authority reduced AI errors: Strong credibility signals helped AI systems select more reliable sources and reduced hallucinated answers
What to take away from December
December didn’t declare the end of SEO. It defined its next phase.
SEO matured into visibility management for AI-driven systems. Success depended on clarity, credibility, and structure, not shortcuts or volume. The fundamentals still worked, but only when applied with discipline.
By the end of 2025, the direction was clear. SEO didn’t get smaller. It got more precise.
SEO evolved into visibility management for AI-driven search. Precision replaced volume.
2025 didn’t rewrite SEO. It clarified it.
Search moved from ranking pages to retrieving answers. From rewarding volume to rewarding clarity. From clicks to credibility. And from optimization tricks to systems-level understanding.
The fundamentals still matter. Technical health, helpful content, and strong SEO foundations are non-negotiable. But they are no longer the finish line. What separates visible brands from invisible ones now is how clearly their content can be understood, trusted, and reused by AI-driven search systems.
Going into 2026, the goal isn’t to outsmart search engines. It’s to make your expertise unmistakable. Write for humans, structure for machines, and build authority that holds up even when clicks don’t follow.
SEO didn’t get smaller this year. It got more precise. Stay with us for our 2026 verdict on where search goes next.
http://dubadosolutions.com/wp-content/uploads/2017/05/dubado-logo-1.png00http://dubadosolutions.com/wp-content/uploads/2017/05/dubado-logo-1.png2025-12-24 12:52:512025-12-24 12:52:51The 2025 SEO wrap-up: What we learned about search, content, and trust
Most business owners assume that if an ad is approved by Google or Meta, it is safe.
The thinking is simple: trillion-dollar platforms with sophisticated compliance systems would not allow ads that expose advertisers to legal risk.
That assumption is wrong, and it is one of the most dangerous mistakes an advertiser can make.
The digital advertising market operates on a legal double standard.
A federal law known as Section 230 shields platforms from liability for third-party content, while strict liability places responsibility squarely on the advertiser.
Even agencies have a built-in defense. They can argue that they relied on your data or instructions. You can’t.
In this system, you are operating in a hostile environment.
The landlord (the platform) is immune.
Bad tenants (scammers) inflate the cost of participation.
And when something goes wrong, regulators come after you, the responsible advertiser, not the platform, and often not even the agency that built the ad.
Here is what you need to know to protect your business.
Note:This article was sparked by a recent LinkedIn post from Vanessa Otero regarding Meta’s revenue from “high-risk” ads. Her insights and comments in the post about the misalignment between platform profit and user safety prompted this in-depth examination of the legal and economic mechanisms that enable such a system.
The core danger: Strict liability explained
While the strict liability standard is specific to U.S. law (FTC), the economic fallout of this system affects anyone buying ads on U.S.-based platforms.
Before we discuss the platforms, it is essential to understand your own legal standing.
In the eyes of the FTC and state regulators, advertisers are generally held to a standard of strict liability.
What this means: If your ad makes a deceptive claim, you are liable. That’s it.
Intent doesn’t matter: You can’t say, “I didn’t mean to mislead anyone.”
Ignorance doesn’t matter: You can’t say, “I didn’t know the claim was false.”
Delegation doesn’t matter: You can’t say, “My agency wrote it,” or “ChatGPT wrote it.”
The law views the business owner as the “principal” beneficiary of the ad.
You have a non-delegable duty to ensure your advertising is truthful.
Even if an agency writes unauthorized copy that violates the law, regulators often fine the business owner first because you are the one profiting from the sale.
You can try to sue your agency later to get your money back, but that is a separate battle you have to fund yourself.
The unfair shield: Why the platform doesn’t care
If you are strictly liable, why doesn’t the platform help you stay compliant? Because they don’t have to.
Section 230 of the Communications Decency Act declares that “interactive computer services” (platforms) are not treated as the publisher of third-party content.
The original intent: This law was passed in 1996 to allow the internet to scale, ensuring that a website wouldn’t be sued every time a user posted a comment. It was designed to protect free speech and innovation.
The modern reality: Today, that shield protects a business model. Courts have ruled that even if platforms profit from illegal content, they are generally not liable unless they actively contribute to creating the illegality.
The consequence: This creates a “moral hazard.” Because the platform faces no legal risk for the content of your ads, it has no financial incentive to build perfect compliance tools. Their moderation AI is built to protect the platform’s brand safety, not your legal safety.
The liability ladder: Where you stand
To understand how exposed you are, look at the legal hierarchy of the three main players in any ad campaign:
The platform (Google/Meta)
Legal status: Immune.
They accept your money to run the ad. Courts have ruled that providing “neutral tools” like keyword suggestions does not make the platform liable for the fraud that ensues.
If the FTC sues, they point to Section 230 and walk away.
The agency (The creator)
Legal status: Negligence standard.
If your agency writes a false ad, they are typically only liable if regulators prove they “knew or should have known” it was false.
They can argue they relied on your product data in good faith.
You (The business owner)
Legal status: Strict liability.
You are the end of the line.
You can’t pass the buck to the platform (immune) or easily to the agency (negligence defense).
If the ad is false, you pay the fine.
The hostile environment: Paying to bid against ‘ghosts’
The situation gets worse.
Because platforms are immune, they allow “high-risk” actors into the auction that legitimate businesses, like yours, have to compete against.
A recent Reuters investigation revealed that Meta internally projected roughly 10% of its ad revenue (approximately $16 billion) would come from “integrity risks”:
Scams.
Frauds.
Banned goods.
Worse, internal documents reveal that when the platform’s AI suspects an ad is a scam (but isn’t “95% certain”), it often fails to ban the advertiser.
Instead, it charges them a “penalty bid,” a premium price to enter the auction.
You are bidding against scammers who have deep illicit profit margins because they don’t ship real products (zero cost of goods sold).
This allows them to bid higher, artificially inflating the cost per click (CPC) for every legitimate business owner.
You are paying a fraud tax just to get your ad seen.
Because the platform is no longer a neutral host but is vouching for the business (“Guaranteed”), regulators can argue they have stepped out from behind the Section 230 shield.
By clicking “Auto-apply,” you are effectively signing a blank check for a robot to write legal promises on your behalf.
Risk reality check: Who actually gets investigated?
While strict liability is the law, enforcement is not random. The FTC and State Attorneys General have limited resources, so they prioritize based on harm and scale.
If you operate in dietary supplements (i.e., “nutra”), fintech (crypto and loans), or business opportunity offers, your risk is extreme. These industries trigger the most consumer complaints and the swiftest investigations.
If you are an HVAC tech or a local florist, you are unlikely to face an FTC probe unless you are engaging in massive fraud (e.g., fake reviews at scale). However, you are still vulnerable to competitor lawsuits and local consumer protection acts.
Investigations rarely start from a random audit. They start from consumer complaints (to the BBB or attorney generals) or viral attention. If your aggressive ad goes viral for the wrong reasons, the regulators will see it.
International intricacies
It is vital to remember that Section 230 is a U.S. anomaly.
If you advertise globally, you’re playing by a different set of rules.
The European Union (DSA): The Digital Services Act forces platforms to mitigate “systemic risks.” If they fail to police scams, they face fines of up to 6% of global turnover.
The United Kingdom (Online Safety Act): The UK creates a “duty of care.” Senior managers at tech companies can face criminal liability for failing to prevent fraud.
Canada (Competition Bureau): Canadian regulators are increasingly aggressive on “drip pricing” and misleading digital claims, without a Section 230 equivalent to shield the platforms.
The “Brussels Effect”: Because platforms want to avoid EU fines, they often apply their strictest global policies to your U.S. account. You may be getting flagged in Texas because of a law written in Belgium.
The advertiser’s survival guide
Knowing the deck is stacked, how do you protect your business?
Adopt a ‘zero trust’ policy
Never hit “publish” on an auto-generated asset without human eyes on it first.
If you use an agency, require them to send you a “substantiation PDF” once a quarter that links every claim in your top ads to a specific piece of proof (e.g., a lab report, a customer review, or a supply chain document).
The substantiation file
For every claim you make (“Fastest shipping,” “Best rated,” “Loses 10lbs”), keep a PDF folder with the proof dated before the ad went live.
This is your only shield against strict liability.
Audit your ‘auto-apply’ settings
Go into your ad accounts today.
Turn off any setting that allows the platform to automatically rewrite your text or generate new assets without your manual review.
Efficiency is not worth the liability.
Watch the legislation
Lawmakers are actively debating the SAFE TECH Act, which would carve out paid advertising from Section 230.
https://i0.wp.com/dubadosolutions.com/wp-content/uploads/2025/12/Why-ad-approval-is-not-legal-protection-Mf1Pt4.webp?fit=1920%2C1080&ssl=110801920http://dubadosolutions.com/wp-content/uploads/2017/05/dubado-logo-1.png2025-12-23 15:00:002025-12-23 15:00:00Why ad approval is not legal protection
Google expanded Demand Gen channel controls to include Google Maps, giving advertisers a new way to reach users with intent-driven placements and far more control over where Demand Gen ads appear.
What’s new. Advertisers can now select Google Maps as a channel within Demand Gen campaigns. The option can be used alongside other channels in a mixed setup or on its own to create Maps-only campaigns.
Why we care. This update unlocks a powerful, location-focused surface inside Demand Gen, allowing advertisers to tailor campaigns to high-intent moments such as local discovery and navigation. It also marks a meaningful step toward finer channel control in what has traditionally been a more automated campaign type.
Response. Advertisers are very excited by this update. CEO of AdSquire Anthony Higman has been waiting for this for decades:
Google Ads Specialist Thomas Eccel, who shared the update on LinkedIn said: “This is very big news and shake up things quite a lot!”
Between the lines. Google continues to respond to advertiser pressure for greater transparency and control, gradually breaking Demand Gen into more modular, selectable distribution channels.
What to watch. How Maps placements perform compared to YouTube, Discover, and Gmail—and whether Google expands reporting or optimization tools specifically for Maps inventory.
First seen. This update was first spotted by Search Marketing Specialist Francesca Poles, when she shared the update on LinkedIn
Bottom line. Adding Google Maps to Demand Gen channel controls is a significant shift that gives advertisers new strategic flexibility and the option to build fully location-centric campaigns.
https://i0.wp.com/dubadosolutions.com/wp-content/uploads/2025/12/Screenshot-2025-12-23-at-14.29.48-8lnVvZ.webp?fit=558%2C447&ssl=1447558http://dubadosolutions.com/wp-content/uploads/2017/05/dubado-logo-1.png2025-12-23 14:56:142025-12-23 14:56:14Google adds Maps to Demand Gen channel controls
Search marketers are starting to build, not just optimize.
Across SEO and PPC teams, vibe coding and AI-powered development tools are shrinking the gap between idea and execution – from weeks of developer queues to hours of hands-on experimentation.
These tools don’t replace developers, but they do let search teams create and test interactive content on their own timelines.
In a zero-click environment, the ability to build unique, useful, conversion-focused tools is becoming one of the most practical ways search marketers can respond.
What is vibe coding?
Vibe coding is a way of building software by directing AI systems through natural language rather than writing most of the code by hand.
Instead of working line by line, the builder focuses on intent – what the tool should do, how it should look, and how it should respond – while the AI handles implementation.
The term was popularized in early 2025 by OpenAI co-founder Andrej Karpathy, who described a loose, exploratory style of building where ideas are tested quickly, and code becomes secondary to outcomes.
His framing captured both the appeal and the risk: AI makes it possible to build functional tools at speed, but it also encourages shortcuts that can lead to fragile or poorly understood systems.
Since then, a growing ecosystem of AI-powered development platforms has made this approach accessible well beyond engineering teams.
Tools like Replit, Lovable, and Cursor allow non-developers to design, deploy, and iterate on web-based tools with minimal setup.
The result is a shift in who gets to build – and how quickly ideas can move from concept to production.
That speed, however, doesn’t remove the need for judgment.
Vibe coding works best when it’s treated as a craft, not a shortcut.
Blindly accepting AI-generated changes, skipping review, or treating tools as disposable experiments creates technical debt just as quickly as it creates momentum.
Mastering vibe coding means learning how to guide, question, and refine what the AI produces – not just “see stuff, say stuff, run stuff.”
This balance between speed and discipline is what makes vibe coding relevant for search marketers, and why it demands more than curiosity to use well.
Vibe coding vs. vibe marketing
Vibe coding should not be confused with vibe marketing.
AI no-code tools used for vibe coding are designed to build things – applications, tools, and interactive experiences.
AI automation platforms used for vibe marketing, such as N8N, Gumloop, and Make, are built to connect tools and systems together.
For example, N8N can be used to automate workflows between products, content, or agents created with Replit.
These automation platforms extend the value of vibe-coded tools by connecting them to systems like WordPress, Slack, HubSpot, and Meta.
Used together, vibe coding and AI automation allow search teams to both build and operationalize what they create.
Why vibe coding matters for search marketing
In the future, AI-powered coding platforms will likely become a default part of the marketing skill set, much like knowing how to use Microsoft Excel is today.
AI won’t take your job – but someone who knows how to use AI might.
We recently interviewed candidates for a director of SEO and AI optimization role.
None of the people we spoke with were actively vibe coding or had used AI-powered development software for SEO or marketing.
That gap was notable.
As more companies add these tools to their technology stacks and ways of working, hands-on experience with them is likely to become increasingly relevant.
Vibe coding lets search marketers quickly build interactive tools that are useful, conversion-focused, and difficult for Google to replicate through AI Overviews or other SERP features.
For paid search, this means teams can rapidly test interactive content ideas and drive traffic to them to evaluate whether they increase leads or sales.
These platforms can also be used to build or enhance scripts, improve workflows, and support other operational needs.
For SEO, vibe coding makes it possible to add meaningful utility to pages and websites, which can increase engagement and encourage users to return.
Returning visitors matter because, according to Google’s AI Mode patent, user state – which includes engagement – plays a significant role in how results are generated in AI Overviews and AI Mode.
For agency founders, CEOs, CFOs, and other group leaders, these tools also make it possible to build custom internal systems to support how their businesses actually operate.
For example, I used Replit to build an internal growth forecasting and management tool.
It allows me to create annual forecasts with assumptions, margins, and P&L modeling to manage the SEO and AI optimization group.
There isn’t off-the-shelf software that fully supports those needs.
Vibe coding tools can also be cost-effective.
In one case, I was quoted $55,000 and a three-month timeline to build an interactive calculator for a client.
Using Replit, I built a more robust version in under a week on a $20-per-month plan.
Beyond efficiency, the most important reason to develop these skills is the ability to teach them.
Helping clients learn how to build and adapt alongside you is increasingly part of the value agencies provide.
In a widely shared LinkedIn post about how agencies should approach AI, Chime CMO Vinneet Mehra argued that agencies and holding companies need to move from “we’ll do it for you” to “we’ll build it with you.”
In-house teams aren’t going away, he wrote, so agencies need to partner with them by offering copilots, playbooks, and embedded pods that help brands become AI-native marketers.
Being early to adopt and understand vibe coding can become a competitive advantage.
Used well, it allows teams to navigate a zero-click search environment while empowering clients and strengthening long-term working relationships – the kind that make agencies harder to replace.
Top vibe coding platforms for search marketers
There are many vibe coding platforms on the market, with new ones continuing to launch as interest grows. Below are several leading options worth exploring.
AI development tool and experience level
Pros
Cons
Google AI Studio (Intermediate)
• Direct access to Google’s latest Gemini models. • Seamless integration with Google ecosystem (Maps, Sheets, etc.). • Free tier available for experimentation.
• Locked into Google’s ecosystem and Gemini models. • Limited flexibility compared to open platforms. • Smaller community/resources compared to established tools.
• Relatively new platform with less maturity. • Limited customization for complex applications. • Generated code may need refinement for production.
Figma Make (Intermediate)
• Seamless design to code workflow within. • Ideal for teams already using Figma. • Bridges gap between designers and developers.
• Requires Figma subscription and ecosystem. • Newer tool, still evolving features. • Code output may need developer review for production.
Replit (Intermediate)
• All-in-one platform (code, deploy, host). • Strong integration capabilities with third-party tools. • No local setup required.
• Performance can lag compared to local development. • Free tier has significant limitations. • Fees can add up based on usage.
Cursor (Advanced)
• Powerful AI assistance for experienced developers. • Works locally with your existing workflow. • Advanced code understanding and generation.
• Steeper learning curve, requires coding knowledge. • Need to download the software GitHub dependency for some features.
For beginners:
Lovable is the most user-friendly option for those with little coding experience.
Figma Make is also intuitive and works well for teams already using Figma.
Replit is also relatively easy to use and does not require prior coding experience.
For developers, Replit and Cursor offer deeper tooling and are better suited for integrations with other systems, such as CRMs and CMS platforms.
Google AI Studio is broader in scope and offers direct connections to Google products, including Google Maps and Gemini, making it useful for teams working within Google’s ecosystem.
You should test several of these tools to find the one that best fits your needs.
I prefer Replit, but I will be using Figma Make because our creative teams already work in Figma.
Bubble is also worth exploring if you are new to coding, while Windsurf may be a better fit for more advanced users.
Practical SEO and PPC applications: What you can build today
There is no shortage of things you can build with vibe coding platforms.
The more important question is what interactive content you should build – tools that do not already exist, solve a real problem, and give users a reason to return.
Conversion focus matters, but usefulness comes first.
Common use cases include:
Lead generation tools
Interactive calculators, such as ROI estimators and cost analyzers.
Quiz funnels with email capture.
Free tools, including word counters and SEO analyzers
Content optimization tools
Keyword density checkers.
Readability analyzers.
Meta title and description generators
Conversion rate optimization
Product recommenders.
Personalization engines.
Data analysis and reporting
Custom analytics dashboards.
Rank tracking visualizations.
Competitor analysis scrapers, with appropriate ethical considerations.
Articles can only take you so far in a zero-click environment, where AI Overviews increasingly provide direct answers and absorb traffic.
Interactive content should be an integral part of a modern search and content strategy, particularly for brands seeking to enhance visibility in both traditional and generative search engines, including ChatGPT.
Well-designed tools can earn backlinks, increase time on site, drive repeat visits, and improve engagement signals that are associated with stronger search performance.
For example, we use AI development software as part of the SEO and content strategy for a client serving accounting firms and bookkeeping professionals.
Our research led to the development of an AI-powered accounting ROI calculator designed to help accountants and bookkeeping firms understand the potential return on investment from using AI across different parts of their businesses.
The calculator addresses several core questions:
Why AI adoption matters for their firm.
Where AI can deliver the most impact.
What the expected ROI could be.
It fills a gap where clear answers did not previously exist and represents the kind of experience Google AI Overviews cannot easily replace.
The tool is educational by design.
It explains which tasks can be automated with AI, displays results directly on screen, forecasts a break-even point, and allows users to download a PDF summary of their results.
AI development software has also enabled us to design additional calculators that deliver practical value to the client’s target audience by addressing problems they cannot easily solve elsewhere.
Vibe coding works best when it follows a structured workflow.
The steps below outline a practical process search marketers can use to plan, build, test, and launch interactive tools using AI-powered development platforms.
Step 1: Research and ideation
Run SERP analysis, competitor research, and customer surveys, and use audience research tools such as SparkToro to identify gaps where AI Overviews leave room for interactive tools.
Include sales, PR, legal, compliance, and cybersecurity teams early in the process.
That collaboration is especially important when building tools for clients.
When possible, involve customers or target audiences during research, ideation, and testing.
Step 2: Create your content specification document
Create a content specification document to define what you want to build before you start.
This document should outline functionality, inputs, outputs, and constraints to help guide the vibe coding software and reduce errors.
Include as much training context as possible, such as brand colors, tone of voice, links, PDFs, and reference materials.
The more detail provided upfront, the better the results.
Begin with wireframes and front-end design before building functionality.
Replit prompts for this approach during setup, and it helps reduce rework later.
Getting the design close to final before moving into logic makes it easier to evaluate usability.
Design changes can always be made later.
Step 4: Prompt like a product manager
After submitting the specification document, continue prompting to refine the build.
Ask the AI why it made specific decisions and how changes affect the system.
In practice, targeted questions lead to fewer errors and more predictable outcomes.
Step 5: Deploy and test
Deploy the tool to a test URL to confirm it behaves as expected.
If the tool will be embedded on other sites, test it in those environments as well.
Security configurations can block API calls or integrations depending on the host site.
I encountered this when integrating a Replit build with Klaviyo.
After reviewing the deployment context, the issue was resolved.
Step 6: Update the content specification document
Have the AI update the content specification document to reflect the final version of what was built.
This creates a record of decisions, changes, and requirements and makes future updates or rebuilds easier.
Save this document for reference.
Step 7: Launch
Push the interactive content live using a custom domain or by embedding it on your site.
Plan distribution and promotion alongside the launch.
This is why involving PR, sales, and marketing teams from the beginning of the project matters.
They play a role in ensuring the content reaches the right audience.
The dark side of vibe coding and important watchouts
Vibe coding tools are powerful, but understanding their limitations is just as important as understanding their strengths.
The main risks fall into three areas:
Security and compliance.
Price creep.
Technical debt.
Security and compliance
While impressive, vibe coding tools can introduce security gaps.
AI-generated code does not always follow best practices for API usage, data encryption, authentication, or regulatory requirements such as GDPR or ADA compliance.
Any vibe-coded tool should be reviewed by security, legal, and compliance professionals before launch, especially if it collects user data.
Privacy-by-design principles should also be documented upfront in the content specification document.
These platforms are improving.
For example, some tools now offer automated security scans that flag issues before deployment and suggest fixes.
Even so, human review remains essential.
Price creep
Another common risk is what could be described as the “vibe coding hangover.”
A tool that starts as a quick experiment can quietly become business-critical, while costs scale alongside usage.
Monthly subscriptions that appear inexpensive at first can grow rapidly as traffic increases, databases expand, or additional API calls are required.
In some cases, self-hosting a vibe-coded project makes more sense than relying on platform-hosted infrastructure.
Hosting independently can help control costs by avoiding per-use or per-visit charges.
Technical debt
Vibe coding can also create technical debt.
Tools can break unexpectedly, leaving teams staring at code they no longer fully understand – a risk Karpathy highlighted in his original description of the approach.
This is why “Accept all” should never be the default.
Reviewing AI explanations, asking why changes were made, and understanding tradeoffs are critical habits.
Most platforms provide detailed change logs, version history, and rollback options, which makes it possible to recover when something breaks.
Updating the content specification document at major milestones also helps maintain clarity as projects evolve.
Vibe coding is your competitive edge
AI Overviews and zero-click search are changing how value is created in search.
Traffic is not returning to past norms, and competing on content alone is becoming less reliable.
The advantage increasingly goes to teams that build interactive experiences Google cannot easily replicate – tools that require user input and deliver specific, useful outcomes.
Vibe coding makes that possible.
The approach matters: start with research and a clear specification, design before functionality, prompt with intent, and iterate with discipline.
Speed without structure creates risk, which is why understanding what the AI builds is as important as shipping quickly.
The tools are accessible. Lovable lowers the barrier to entry, Cursor supports advanced workflows, and Replit offers flexibility across use cases.
Many platforms are free to start. The real cost is not testing what’s possible.
More importantly, vibe coding shifts how teams work together.
Agencies and in-house teams are moving from “we’ll do it for you” to “we’ll build it with you.”
Teams that develop this capability can adapt to a zero-click search environment while building stronger, more durable partnerships.
Build something. Learn from it. The competitive advantage is often one prompt away.
Brand-agency partnerships look very different today than they did even a few years ago, and by 2026 that gap will only widen.
Internal marketing teams are more sophisticated, digital channels are more specialized, and the role agencies play is no longer one-size-fits-all.
As a result, the companies that get the most value from agency relationships aren’t always the biggest spenders.
They’re the ones that are clear about what they need and what they don’t.
That clarity starts with understanding the true role an agency should play inside your organization.
Too many partnerships struggle because expectations and responsibilities were never properly aligned from the start.
When that foundation is off, even strong execution can fall flat.
After working with thousands of businesses across various industries and growth stages, we consistently observe that agency success falls into two distinct partnership models, primarily shaped by company size and internal marketing maturity.
Model 1: Execution-first partnerships (large companies)
If your company generates more than $50 million in annual online revenue, you likely already have a strong internal marketing team.
Strategy, goal-setting, and planning live in-house. What you need from an agency is deep platform expertise and consistent, high-level execution.
At this stage, agencies function as specialist operators that:
Activate the roadmap your team has already defined.
Optimize performance inside specific channels.
Bring advanced technical knowledge that would be inefficient to replicate internally.
When something underperforms, a strong agency partner doesn’t rush to tactics.
They help determine whether the issue lies in execution, shifting market conditions, or a broader strategic blind spot – and they bring the data needed to support course correction.
Model 2: Integrated growth partners (small to mid-size companies)
For companies under $50 million in annual online revenue, the agency relationship is different.
Internal teams are often lean, stretched, or still developing core digital expertise.
In these cases, agencies don’t just execute – they help shape the entire growth strategy.
Here, the right agency partner becomes an extension of the marketing department that can:
Guide platform selection.
Develop cross-channel strategies.
Execute campaigns.
Provide direction on tools, tracking, and infrastructure.
The relationship is more integrated because it has to be.
For many growing businesses, agencies offer access to senior-level expertise at a fraction of the cost of building a full in-house team.
That tradeoff often creates the best possible balance between speed, strategy, and financial reality.
Most companies approach agency selection the wrong way.
Here’s how to improve your odds of finding a partner that actually fits your needs.
Ditch the RFPs
Many large companies use the request for proposal (RFP) process to solicit potential partners.
However, RFPs often favor vendors that excel at paperwork over those that prioritize performance.
From an agency perspective, if you don’t already know you’ve won an RFP, you’re not going to win it.
They act more as rubber stamps for a decision that has already been made.
Large companies should instead leverage their connections.
If you’re running a large internal marketing department, you probably already know dozens of professionals who could provide referrals.
Use that network to find firms doing great work, then reach out to them directly.
Smaller businesses should talk to their peers about trusted marketing vendors and then check reviews to validate those recommendations.
No agency is perfect, and every agency will have some dissatisfied clients.
But if you see patterns of negative reviews emerge, you should stay away.
Request an audit
Once you’ve identified a few potential partners, ask them to audit your current marketing setup.
In most cases, digital marketing agencies conduct these audits for free.
Keep in mind that during an audit, many agencies will point out what you’re doing wrong.
But the goal is to receive honest, constructive feedback that offers insight into what’s working and what’s possible.
The audit process will look different depending on the company’s size.
For larger companies, agencies should only audit the platforms they’ll be working on.
Smaller companies need a broader audit across the entire marketing funnel.
These agencies won’t be working in a vacuum.
Every element of marketing is interrelated, so they’ll need to know who manages each stage of the funnel and whether they’re doing a good job.
Companies of all sizes should collect audits from multiple sources.
This enables you to compare recommendations and understand if the partnership will be a good fit.
Large companies need partners that can integrate with their internal processes.
Smaller companies need to pick vendors with people they actually want to work with.
Both considerations are critical in ensuring long-term success.
Setting achievable goals
Once you’ve selected the right agency partner, it’s time to define your goals.
It’s an unfortunate reality that most business leaders set marketing goals that don’t align with their business goals, which puts agency partners in an untenable position before the relationship even gets off the ground.
Good agencies should challenge your goals before you even sign a contract. They should push you to dream bigger or rein you in if your expectations are unrealistic.
If a potential client in the beauty space says they want a tenfold return on ad spend (ROAS) while jumping their non-brand spend from $20,000 to $100,000, a good agency should know enough to push back.
Your potential partner should understand the economics of your business and help ensure your marketing goals align with your business goals.
Often they don’t, which is where good agencies add immediate value.
Once the work begins, you need to keep your agency accountable. Here’s how.
Contract length
Larger companies typically sign 12-month contracts with their agency vendors.
They value stability and performance, and longer contract terms provide agencies with the time needed to establish themselves within the marketing operation.
Smaller companies can’t afford to bind themselves to an underperforming agency for an entire calendar year.
If you’re hiring an agency partner at a smaller company, opt instead for a three-month agreement that automatically renews to month-to-month.
Challenge and conflict are healthy
The most productive business-agency partnership often involves some conflict from time to time.
Great partners will challenge your thinking regularly, which can sometimes create discomfort.
But if everything is always smooth sailing, you probably aren’t growing or improving.
The goal instead is to have productive conversations that involve healthy disagreement and constant refinement.
Ongoing accountability
If you’re overseeing a brand-agency partnership, you should establish regular reviews that compare progress to the opportunities identified in the agency’s initial audit.
For smaller companies, quarterly reviews make sense. They align with the contract structure and allow you to recalibrate budget allocation.
Larger companies might review monthly or quarterly, depending on spend and complexity.
However, context here matters. You need to understand if your industry is growing or shrinking to judge your agency’s work.
For example, if your industry is down 10% year-over-year and your sales are flat, you’re outperforming your competitors.
Often, the agency or brand can obtain this information from their representatives on platforms such as Google, Microsoft, Amazon, or Meta.
Innovation and testing
Great agency partners will proactively bring new growth ideas to the table, which is particularly valuable for smaller businesses.
Large companies also benefit from outside ideas and should establish dedicated budgets for testing.
After all, if your agency isn’t investing at least a small portion of the budget into new, untested ideas, brands will find themselves falling behind competitors that are.
Innovation isn’t just about testing what works today. It’s about understanding what’s coming next.
Great agency partners should help you see what’s coming 6-12 months out, and prepare your marketing to meet those new conditions.
Businesses need an agency’s expertise, which becomes insight over the longer term.
Not every brand-agency partnership succeeds, even with the best intentions.
If your gut is telling you something isn’t working or that something could be working better, here are a few red flags that might indicate it’s time to make a change.
Your business isn’t growing
Your marketing efforts should revolve around finding new-to-brand customers. Full stop.
If your business isn’t growing and your industry is stable or growing, that’s a big red flag that marketing isn’t working.
Once an agency stops being a partner in growth, it’s time to make a change.
Your agency isn’t pushing innovation
The marketing ecosystem is constantly changing:
Customer needs evolve.
Platforms update features.
New tools emerge that upend old processes.
If your agency isn’t bringing new ideas or exploring new ways to reach customers, your marketing is stagnating.
In these instances, an outside audit can reveal deficiencies and potential opportunities.
Your agency can’t explain performance
If your agency can’t contextualize your performance – good or bad – within the broader marketing ecosystem, it’s a strong indication they don’t understand your sales funnel.
Channel experts should know how their performance is affected by upper-funnel activities and how those activities affect bottom-funnel activities.
Marketing agencies for smaller businesses should know enough about the entire marketing operation and understand how performance in one area impacts another.
https://i0.wp.com/dubadosolutions.com/wp-content/uploads/2025/12/What-successful-brand-agency-partnerships-look-like-in-2026-gOVqTH.webp?fit=1920%2C1080&ssl=110801920http://dubadosolutions.com/wp-content/uploads/2017/05/dubado-logo-1.png2025-12-23 13:00:002025-12-23 13:00:00What successful brand-agency partnerships look like in 2026
Generative AI is everywhere right now. It dominates conference agendas, fills LinkedIn feeds, and is reshaping how many businesses think about organic search.
Brands are racing to optimize for AI Overviews, build vector embeddings, map semantic clusters, and rework content models around LLMs.
What gets far less attention is a basic reality: for most websites, AI platforms still drive a small share of overall traffic.
AI search is growing, no question.
But in most cases, total referral sessions from all LLM platforms combined amount to only about 2% to 3% of the organic traffic Google alone delivers.
Despite that gap, many teams are spending more time chasing AI strategies than fixing simple, high-impact SEO fundamentals that continue to drive measurable results.
Instead of improving what matters most today, they are overinvesting in the future while underperforming in the present.
This article examines how a narrow focus on AI can obscure proven SEO tactics and highlights practical examples and real-world data showing how those fundamentals still move the needle today.
1. Quick SEO wins are still delivering outsized gains
In an era where everyone is obsessed with things like vector embeddings and semantic relationships, it’s easy to forget that small updates can have a big impact.
For example, title tags are still one of the simplest and most effective SEO levers to pull.
And they are often one of the on-page elements that most websites get wrong, either by targeting the wrong keywords, not including variations, or targeting nothing at all.
Just a few weeks ago, a client saw a win by simply adding “& [keyword]” to the existing title tag on their homepage. Nothing else was changed.
Keyword rankings shot up, as did clicks and impressions for queries containing that keyword.
This was all achieved simply by changing the title tag on one page.
Couple that with other tactics, such as on-page copy edits, internal linking, and backlinking across multiple pages, and growth will continue.
It may seem basic, but it still works.
And if you only focus on advanced GEO strategies, you may overlook simple tactics that provide immediate, observable impact.
2. Content freshness and authority still matter for competitive keywords
Another tactic that has faded from view with the rise of AI is what’s often called the skyscraper technique.
It involves identifying a set of keywords and the pages that already rank for them, then publishing a materially stronger version designed to outperform the existing results.
It’s true that the web is saturated with content on similar topics, especially for keywords visible in most research tools.
But when a site has sufficient authority, a clear right to win, and content freshness, this approach can still be highly effective.
I’ve seen this work repeatedly.
Here’s Google Search Console data from a recent article we published for a client on a popular, long-standing topic with many competing pages already ranking.
The post climbed to No. 2 almost immediately and began generating net-new clicks and impressions.
Why did it work?
The site has strong authority, and much of the content ranking ahead of it was outdated and stale.
If you’re hesitant to publish the thousandth article on an established topic, that hesitation is understandable.
This approach won’t work for every site. But ignoring it entirely can mean passing up clear, high-confidence wins like these.
3. User experience remains a critical conversion lever
Hype around AI-driven shopping experiences has led some teams to believe traditional website optimization is becoming obsolete.
There is a growing assumption that AI assistants will soon handle most interactions or that users will convert directly within AI platforms without ever reaching a website.
Some of that future is beginning to take shape, particularly for ecommerce brands experimenting with features like Instant Checkout in ChatGPT.
But many websites are not selling products.
And even for those that are, most brands still receive a significant volume of traffic from traditional search and continue to rely on calls to action and on-page signals to drive conversions.
It also makes little difference how a user arrives – via organic search, paid search, AI referrals, or direct visits.
A fast site, a strong user experience, and a clear conversion funnel remain essential.
There are also clear performance gains tied to optimizing these elements.
Here are the results we recently achieved for a client following a simple CTR test:
Brands that continue to invest in user experience and conversion rate optimization will outperform those that do not.
That gap is likely to widen the longer teams wait for AI to fully replace the conversion funnel.
AI is reshaping search, but what works still matters
There is no dispute that AI is reshaping the search landscape.
It’s changing user behavior, influencing SERPs, and complicating attribution models.
The bigger risk for many businesses, however, is not underestimating AI but overcorrecting for it.
Traditional organic search remains the primary traffic source for most websites, and SEO fundamentals still deliver when executed well.
Quick wins are real.
Higher-quality content continues to be rewarded.
User experience optimization shows no signs of becoming irrelevant.
These are just a few examples of tactics that remain effective today.
Importantly, these efforts do not operate in isolation.
https://i0.wp.com/dubadosolutions.com/wp-content/uploads/2025/12/AI-referral-sessions-vs-Google-organic-clicks-3cpFkz.webp?fit=1128%2C438&ssl=14381128http://dubadosolutions.com/wp-content/uploads/2017/05/dubado-logo-1.png2025-12-22 18:13:042025-12-22 18:13:04AI search is growing, but SEO fundamentals still drive most traffic
AI is reshaping how people shop online. Search isn’t just about keywords anymore. Tools like Google’s AI Overviews, ChatGPT shopping features, and Perplexity product recommendations analyze huge amounts of product data to decide what to show users. That shift means e-commerce brands need to rethink the way their product information is structured.
If you want visibility in these AI-powered shopping journeys, your product data has to be clean, complete, and enriched. AI models lean heavily on structured feeds, trusted marketplaces, and high-quality product attributes to understand exactly what you sell.
That’s why AIsearch for e-commerce matters right now. Brands that optimize their feeds will show up in conversational queries, comparison results, and visual search responses. Brands that don’t will struggle to appear even if they’ve done traditional SEO well.
This foundation will help you give AI systems the clarity they need to recommend your products with confidence.
Key Takeaways
AI search engines rely heavily on structured product feed data instead of just site content to understand and surface products.
Clean, complete feeds lead to higher visibility across Google Shopping, ChatGPT shopping research, Perplexity results, and other LLMs.
Strong titles, enriched attributes, and quality images make it easier for AI systems to match your products to real user needs.
Brands with clear, structured product data will outperform competitors in AI-driven shopping experiences.
How AI Search Is Reshaping Product Discovery
AI is changing the way customers find products long before they reach your website. Instead of typing traditional keywords, shoppers now describe what they want in plain language: “lightweight waterproof hiking boots,” “a gift for a 12-year-old who loves science,” “a mid-century floor lamp under $150.”
AI systems interpret these natural-language queries using semantic understanding instead of exact keyword matches. That shift affects everything from Google Shopping listings to ChatGPT’s built-in shopping tools. It also impacts how AI-driven platforms rank your products when answering conversational or comparison-based queries.
If you’ve been following the evolution of AI in e-commerce, you already know AI is moving deeper into product search, recommendation, and personalization. But behind the scenes, the link between your product data and AI visibility is tightening.
AI models rely on structured, trustworthy data sources, including product feeds, schema markup, and marketplace listings. If your feed lacks attributes or clarity, AI can’t confidently connect your product to a user’s need, even if your website is strong.
Optimizing your feed is no longer a backend task. It’s a visibility strategy.
What Is a Product Feed (and Why AI Cares About It)
A product feed is a structured data file that contains detailed information about every item you sell. It includes attributes like product title, description, brand, size, color, price, availability, GTIN, and more. Platforms such as Google Shopping, Meta, Amazon, and TikTok Shops rely on these feeds to understand your inventory and decide when to show your products.
AI systems depend on the same structure. Instead of scanning pages manually, they pull product details from feeds because the information is cleaner, more complete, and easier to interpret at scale.
If your feed includes rich attributes, AI can match your items to complex user queries. When attributes are missing or titles are vague, your products become invisible in AI-driven discovery, regardless of how strong your website content might be.
This is why optimizing product feeds is a priority for e-commerce brands right now. Clean, enriched feeds increase your visibility across AI-powered shopping experiences and visual search tools like Google Lens.
Now, your product feed isn’t just for ads, but is a core input for AI search.
What AI Needs From Your Product Feed (Titles, Attributes, Images)
AI systems don’t guess what your products are, instead analyzing the data you provide. These are the elements that matter most.
Titles and Descriptions
AI models prefer natural, descriptive, human-sounding titles. Short, vague titles like “Running Shoes” don’t give AI enough context. But a title such as:
instantly signals the audience, category, and key benefits.
Descriptions should reinforce the title and add details that help AI understand use cases, materials, fit, and core value.
Avoid keyword stuffing. AI systems would likely reference sites with ambiguity less because they would have less info to understand it.
Product Attributes
AI engines rely heavily on structured attributes such as:
Size
Color
Material
Fit
Style
GTIN/MPN
Age range
Intended use
Missing attributes = missing visibility.
Attributes help AI refine products when users ask things like: “Show me a size 8,” “Only vegan options,” “Something in walnut or dark wood.”
The more complete your attributes, the better your likelihood of appearing in those filtered results.
Product Images and Alt Text
AI increasingly “reads” images using vision models. Google Lens, Pinterest Lens, and multimodal AI systems analyze colors, textures, shapes, and packaging.
Clear, high-resolution images paired with alt text provide two inputs: visual interpretation and descriptive language.
Example alt text: “Women’s waterproof trail running shoe with rubber sole, breathable mesh upper, and reinforced toe cap in blue.”
Visual clarity improves both AI understanding and user experience.
Steps To Optimize Product Feeds for AI Visibility
Here’s the practical workflow to upgrade your product feed for AI search visibility.
1. Audit Your Current Product Feed
Start with a complete audit using tools like Google Merchant Center, Feedonomics, or GoDataFeed. Look for:
Missing GTINs or invalid identifiers
Weak or vague product titles
Incomplete attributes
Duplicate listings
Mismatched availability or pricing
Blank fields or generic descriptions
AI search systems penalize incomplete or ambiguous data.
Include product schema markup on all product pages, especially:
Product
Review
Price
Availability
AI search engines treat structured schema as a trust signal.
Also include descriptive alt text on all product images to support accessibility and AI interpretation.
5. Set Up Feed Rules and Automations
Automate cleanup tasks such as:
Adding missing colors to titles
Appending product type or material
Standardizing capitalization
Populating missing attributes with known defaults
Flagging products with incomplete data
Automation keeps your feed consistent as your catalog changes.
How AI Assistants Use Product Data
AI shopping assistants are rapidly changing how customers discover and compare products.
To generate these answers, AI systems pull from:
Merchant Center feeds
Structured schema markup
Marketplace listings
Verified product databases
High-quality product images
Trusted review sources
This creates a composite understanding of your product beyond just what your site says about it.
If you’ve explored the role of AI shopping assistants, you’ve likely seen how quickly they recommend products based on attributes like size, color, performance, ratings, and price. Those signals come directly from your feed and structured product data.
Brands with richer data sets see higher inclusion rates in:
AI systems don’t guess. They promote products they can understand clearly and ignore the rest.
Common Mistakes That Hurt AI Visibility
Most feed problems fall into a few categories, and each one reduces visibility in AI search engines.
1. Vague or Duplicated Titles
Titles like “Running Shoes” or “LED Lamp” provide no usable context. AI deprioritizes these compared to richer alternatives.
2. Missing Key Attributes
Many merchants skip fields like size, color, material, GTIN, or gender. AI relies heavily on these attributes when matching products to specific user requests.
3. Keyword-Stuffed or Fluffy Descriptions
Descriptions should be informative, not bloated. AI models prefer specific phrasing over repetitive keywords.
4. Inconsistent Pricing or Availability
If your feed shows “in stock” but your page says “out of stock,” AI systems flag inconsistencies and may reduce your visibility.
5. Low-Quality Images or Missing Alt Text
Visual AI models need clarity. Poor images or missing alt text make your product harder to classify.
Fixing these issues has a measurable impact on how often your products appear in AI-driven recommendations.
FAQs
What is AI e-commerce?
AI e-commerce refers to using artificial intelligence to improve product discovery, recommendations, personalization, and automation throughout the online shopping experience.
How is AI changing e-commerce?
AI is shifting product discovery toward natural-language search, visual identification, and conversational shopping assistants. Brands now need structured, enriched product data to stay visible.
How do you optimize a product feed for AI search?
Create clear titles, use complete attributes, include schema markup, strengthen product images, and use automation to maintain consistency. A detailed feed helps AI understand your products accurately.
Conclusion
Brands that invest in structured data, enriched attributes, and clear product information will outperform competitors as AI-driven shopping grows.
Feed optimization also strengthens your broader search strategy. The same structured data powering AI engines aligns with strong AI in e-commerce practices, and the same clarity helps conversational systems recommend your products more confidently.
Visibility in AI search isn’t random. It comes from data quality. And improving that data is one of the highest-impact steps an e-commerce brand can take today.
