After The AI Content Flood: Why Personalized Experiences Need Better Content Architecture

Scaling AI Content Is The #1 Enterprise Priority: How Do You Scale Without Penalty?

Enterprise content leaders are scaling AI output and struggling to do it well. The highest-maturity organizations already know why.

Shelley Walsh 3.3K Reads

Scaling AI Content Is The #1 Enterprise Priority: How Do You Scale Without Penalty?

Scaling AI content generation is the number one content strategy for enterprise organizations optimizing for AI search visibility. According to Conductor’s 2026 State of AEO/GEO CMO Investment Report, which surveyed over 250 executives and digital leaders across 12 industries, it ranked above structured data, above authoritative long-form guides, and above original research. Across every maturity level surveyed, from organizations venturing into AI visibility to those with enterprise-wide adoption, it was the top answer.

However, this may also be where the problem starts.

AI Content Scaling Is Failing

Inside the report, Aleyda Solis acknowledged the strategic intent but raised a concern: “Although it’s possible to leverage AI for content, a personalized editorial and optimization workflow is required to ensure quality, originality, and expertise by integrating unique brand insights and first-party data, which is exactly what AI platforms are likely to cite.”

Eli Schwartz predicted that the current AI content scaling trend “will change in 2026 as Google and other LLMs push back against low-quality content” with what he described as an AI version of Google’s Helpful Content Update. He also flagged that the leaders he speaks with are “somewhat skeptical about the effectiveness of mass amounts of AI content, but are afraid of being left behind if they don’t do this.”

Fear of missing out is not a basis for an effective content strategy.

Lily Ray, who is known for her in-depth analysis, said earlier this year: “Interesting, but not surprising, to see people on LinkedIn sharing their stories of losing all search visibility (sometimes overnight) after an aggressive AI content strategy.” She added: “Just because it’s easy doesn’t mean it’s a good idea.”

I strongly echo that if something is easy, it’s easy for everyone and not competitive.

Pedro Dias documented that in June 2025, Google began issuing manual actions specifically for scaled content abuse, targeting sites that had been mass-publishing AI-generated content. Sites across the UK, US, and EU received Search Console notifications citing “aggressive spam techniques, such as large-scale content abuse.”

Dan Taylor recently wrote about the mechanics of this failure in granular detail, sharing traffic graphs that illustrate what Glenn Gabe calls the “Mt. AI” effect, an initial spike when new content floods the index, followed by a cliff edge as Google’s quality threshold assessment kicks in. What Taylor identifies as the real problem isn’t AI content itself, but the absence of any genuine content strategy underneath it. “The real problem lies in the fact that scaling content production, regardless of the method, often introduces a raft of quality control issues,” he writes. The freshness boost that new URLs receive masks those issues temporarily. Then it doesn’t.

I write, read, and edit a lot of content, and I can clearly see when AI has been used to supplement writing. Some writers can do this well and have input enough of their expertise to get reasonable results. Others not so much, where they are leaning on AI to supplement their lack of knowledge or expertise. For myself, I can get astounding results from Claude when I input quality, unique research, but I do have to invest a huge amount of guidance to get anything worth publishing.

To be clear, I’m not anti-AI usage. Like Google, I’m focused on good quality content and writing.

That gap between what AI produces by default and what’s actually publishable is precisely where the opportunity still lives for writers who know their subject. Exceptional human-guided content isn’t a compromise. Right now, it’s the competitive advantage.

Google Is Consistent About AI Content

Google’s position on the use of AI content and quality content has been consistent.

Danny Sullivan spoke at the Google Search Central event in Toronto in April 2026 about the concept of commodity versus non-commodity content.

Commodity content is everything an AI can produce from publicly available information. Non-commodity content requires you to have actually done something, know something from direct experience, or hold an opinion grounded in genuine expertise. And this is what Google considers your competitive strength going into the AI era.

John Mueller framed AI content abuse in the context of Google’s Quality Rater Guidelines update, which now explicitly groups AI-generated content in a section about content created with little effort or originality. Quality raters are instructed to apply the lowest rating to pages where all or almost all of the content is auto- or AI-generated with little to no effort, originality, or added value, regardless of production method. Google’s guidelines are explicit that AI tools alone don’t determine the rating, effort, originality, and value do.

This all aligns with the foundations of what Google wants to surface – quality content that demonstrates first-hand experience.

We Have Seen This Before

Lily Ray ran a test by asking Perplexity for SEO news and received a confident report about the “September 2025 Perspective Core Algorithm Update,” a Google update that had never happened. The citations Perplexity provided pointed to AI-generated posts on SEO agency blogs. Sites that had run a content pipeline, hallucinated an update, and published it as reporting. Perplexity read this and treated it as source material, and served it back to her as fact.

There’s a historical parallel here that some older SEOs will recognize.

Early digital PR/link building efforts involved seeding stories or content into lower-tier publications because top-tier journalists used them as source material, and it generated implied credibility of multiple citations. Journalists then began to cite what was published by other sites, and published sites cited and referenced them in the same citation cycle.

Another example I saw recently involved several articles [incorrectly] reporting that Jeremy Clarkson and his partner Lisa Hogan (from the top Amazon UK show Clarkson’s Farm) were spending time apart and ending their relationship. What Clarkson had actually said was that they deliberately go their separate ways during the day so they have something interesting to talk about in the evening. This might be a low-stakes example, but it perfectly illustrates how quickly misinformation spirals.

Screenshot from search for [have jeremy clarkson and lisa hogan split up], Google UK, May 2026

Content Scale Is Strategy And Challenge

The highest-maturity organizations in the Conductor report (organizations where AEO/GEO is a core digital priority) have already arrived at the right conclusion, and they are the only group in the study that prioritized original research based on first-party data as a content strategy. They understand that first-party data and genuine research cannot be replicated by running an AI content operation and exclusivity is the point.

The Conductor report’s headline finding is that 94% of enterprise organizations plan to increase AEO/GEO investment in 2026, and that AEO/GEO has become the number one marketing priority, above paid media and paid search. The report also surfaces that generating AI-optimized content at scale is not only the top stated strategy, but also the top stated challenge. Brands know what they want to do, but they don’t know how to get there.

How Enterprise Brands Can Scale And Win

Industries that already operate on programmatic content models (travel, ecommerce, large product catalog sites) have been producing content at scale for years. A hotel comparison site generating location pages, a retailer producing thousands of product descriptions, a marketplace creating structured listings are all legitimate use cases where AI can effectively accelerate something that was already happening.

But, to have real brand differentiation, investing in a unique voice and approach to how they write these listings can set them apart and be a competitive advantage.

Alongside their programmatic content, enterprise brands should also be finding ways they can produce content that is genuinely difficult to replicate. Experience-driven, data-grounded, editorially considered, and specific in ways that only a real subject matter expert would know.

For an enterprise brand to win at scaling content, my recommendation is to wrap AI usage around subject-matter experts and editors. The power of AI is how it can turn experts into super producers and allow them to produce more. Enterprise brands should invest in finding these super producers and then use AI to exponentially scale their ability, not try and replace them.

AI Amplifies What’s Already There

The most useful frame for AI in content production is as an amplifier of whatever you bring to it. If you have genuine subject matter knowledge, proprietary data, and the editorial discipline to maintain quality, AI can meaningfully accelerate your output. It helps you produce more of what you’re already good at, faster.

But if you don’t have those things, AI produces more of what you don’t have, faster. The content output has structure, length, and the right vocabulary, but it contains nothing that an LLM can’t generate from publicly available information. Nothing that differentiates you from every other brand trying to scale with AI in the same way.

As I said earlier, I have produced in-depth content for years, and for me, AI is a creative amplifier and an exciting tool that augments what I know. It doesn’t replace me, and it certainly can’t do what I can by itself. On that basis, I see subject-expert editors as being the new information gatekeepers.

For enterprise brands who want to scale their content they should start with understanding that good content is not about including everything; it’s about knowing what not to include.

The full Conductor 2026 State of AEO/GEO CMO Investment Report is available here.

More Resources:

Featured Image: ImageFlow/Shutterstock

From Personalization to Performance: Why Marketing Leaders Need a New Content Architecture

contentful

From Personalization to Performance: Why Marketing Leaders Need a New Content Architecture

We’re producing more content today than at any point in human history, yet audiences have never been harder to reach. The two facts aren’t coincidental; they’re causally linked. It’s cheaper than ever to produce content — and when every brand can publish at volume, it’s no longer an advantage, it’s just noise.

If you’ve been watching your conversion rates flatten and your time-on-site erode, the reflex to personalize is the correct one. The difficulty is in how that reflex translates into work. Most personalization programs are a series of isolated optimizations: a landing page tuned for one audience, a headline variant tested, a win declared, attention shifted to the next surface. Each effort starts from zero. This creates a string of disconnected experiments rather than what your brand really needs: a continuous experience that remembers who your visitor is, what they care about, and reflects that memory across every touchpoint they have with you.

The brands that approach personalization this way build experiences that pull customers in and hold their attention across the entire buyer’s journey. What gets them there isn’t a new tool, or a better model, or a bigger experimentation budget. It’s a structural decision about how content is organized underneath everything else.

The architecture that’s holding you back

Here’s the scenario: You work for a car manufacturer and your marketing team identifies a high-value segment – someone looking for a luxury car. You decide to tailor the homepage hero image, product page messaging, and email nurture for that audience. It’s a sound idea and a defensible priority. A ticket goes to engineering.

Three weeks later, what comes back is a duplicated page with hardcoded copy for one segment. The nurture stream is still generic because it lives in a different system entirely. The “personalization initiative” produced exactly one surface, consumed a dev sprint, and generated enough maintenance overhead to ensure that nobody on the team wants to attempt anything similar for at least six months.

The problem is structural rather than tactical. When a CMS treats content as pages, which is to say as monolithic units where layout, copy, and presentation are fused together, personalization can only ever operate at the page level. The instrument is blunt by definition: this audience sees page A, that audience sees page B. The granularity that actually moves performance requires content broken into components that can be assembled by context rather than hardcoded by a developer.

Many teams reach for a script tag and a third-party personalization tool. The appeal? Marketers can change content without a developer. But it comes at a cost. Those scripts introduce latency and shift layout, they cause visible flicker and degrade Core Web Vitals. And because they operate outside the CMS, they erode whatever governance and guardrails the team has spent years building.

Modular, composable content architecture solves the page-level problem and the script-tag problem in the same move. Personalization happens inside the content system itself, inheriting the same governance, workflows, and roles your team has already invested in. Marketers gain autonomy inside the guardrails defined by your content model and publishing process. The brand team and the engineering team stop dreading the next personalization request, because that request no longer threatens the existing way of working.

This is the architecture that makes it possible to build tailored journeys that compound customer interest over time.

Extending the journey

Most conversations about optimization circle around the question of experimentation velocity: how quickly can you test a variant, read the data, and ship the next iteration? The question has merit, but it keeps the work anchored to individual touchpoints. Even when velocity goes up, what you’re producing is a faster series of isolated optimizations. Each experiment concludes, the next one has to be set up from scratch, and there’s no flywheel of accumulated understanding.

The more interesting question, and the one that separates personalization-as-tactic from personalization-as-capability, is how quickly you can compose a coherent journey that spans multiple surfaces, and how much more specific that journey becomes as you accumulate understanding of a visitor’s context, their intent, and what they’re actually trying to accomplish. That kind of journey does not need to begin everywhere at once. It starts with the surface where context is already easiest to read.

The best starting point is often the page where you are already spending money to send people. Paid search and paid social landing pages are useful for exactly this reason: the audience signal already exists in the ad, keyword, campaign, or UTM parameter, and the outcome is already being measured. If a company is paying to earn a click from a specific promise, the destination should not flatten that context back into generic messaging. The page should continue the conversation the ad began.

But the most important part of paid traffic personalization is what happens after the landing page. When the signal that brought a visitor in, such as the ad creative they clicked, the keyword they searched, or the campaign they came from, carries through to the homepage, the product page, or the next email, the experience stops feeling like a landing page and starts feeling like a brand that understood them from the first moment.

From there, the next useful surface is often the distinction between new and returning visitors on high-traffic organic pages. New and returning visitors are one of the largest natural audience splits most teams already have, and they require no external data to identify. A first-time visitor may need orientation. A returning visitor already has some context and is usually back because something specific is pulling them forward.

That is where the learning starts to compound. What you discover about returning visitor behavior, including which content pulls them deeper, which CTAs convert, and which messages land, becomes the foundation for more refined audience definitions later. The first experiment answers the obvious question; the data from that experiment tells you what the next, more specific question should be.

From campaigns to continuous capability

Most teams still treat personalization as a project: a kickoff, a roadmap, a defined set of experiences to build, a launch date. Six months later, those experiences are still running unchanged because nobody has the bandwidth or the underlying setup to iterate on them. This pattern plays out predictably across teams. Something gets built, launched, and then quietly fails to deliver the value it was scoped for, not because the idea was wrong, but because the program was never set up to sustain itself.

The alternative is treating personalization as an ongoing capability rather than a one-time launch. A capability runs continuously, learns from every interaction, and improves without requiring a new project each time the appetite for personalization returns.

A capability has owners, a regular review rhythm, and a backlog of next tests to run. A project has a launch date. If you can name the person who reviews personalization performance every week, who maintains the test backlog, and who decides what gets retired, then what you have is a capability. If you can’t, what you have is a project that simply hasn’t been recognized as abandoned yet.

AI accelerates this, though not in the way it’s most often described. The real value of AI in personalization isn’t that it generates more variants. It shortens the gap between a visitor’s behavior and a personalized response, which lets your marketers stop being the bottleneck and start spending their time on the creative and strategic decisions that actually require human judgment.

The takeaway

The brands that break through the noise won’t be the ones with the most content, the most sophisticated models, or the biggest experimentation budgets. They’ll be the ones whose experiences are remembered, the ones where every visitor feels recognized rather than processed.

That outcome is architectural before it is strategic, though the choice to pursue it is entirely strategic. It’s a decision to treat your audience as people mid-journey who deserve continuity, rather than as segments to be sorted into buckets. Begin somewhere small, prove the model on a single surface, and then extend the journey forward. What you’ll find is that personalization, done this way, compounds in a manner that isolated testing has never been able to.

For more resources on how composable content architecture supports personalized customer journeys, visit Contentful.com.

Sponsored

Keep it together with Contentful

The go-to DXP for AI, Contentful manages everything from GEO to personalization, removing the roadblocks of modern marketing

Learn more

How Search Engines Tailor Results To Individual Users & How Brands Should Manage It

Turn fragmented SERPs and AI summaries into an advantage by building a brand structure that search engines and users can trust.

Montserrat Cano 3.2K Reads

How Search Engines Tailor Results To Individual Users & How Brands Should Manage It

How many times have you seen different SERP layouts and results across markets?

No two people see the same search results, as per Google’s own documentation. No two users receive identical outputs from AI platforms either, even when using the same prompt. In a time of information overload, this raises an important question for global marketers: How do we manage and leverage personalized search experiences across multiple markets?

Today, clarity and transparency matter more than ever. Users have countless choices and distractions, so they expect experiences that feel relevant, trustworthy, and aligned with their needs in the moment. Personalization is now central to how potential customers discover, evaluate, and engage with brands.

Search engines have been personalizing results for years based on language, search behavior, device type, and technical elements such as hreflang. With the quick evolution of generative artificial intelligence (AI), personalization has expanded into summarized answers on AI platforms and hyper-personalized experiences that depend on internal data flows and processes.

This shift forces marketers to rethink how they measure visibility and business impact. According to McKinsey, 76% of users feel frustrated when experiences are not personalized, which shows how closely relevance and user satisfaction are linked.

At the same time, long-tail discovery increasingly happens outside of search engines, particularly on platforms like TikTok. Statista reports that 78% of global internet users now research brands and products on social media.

All of this is happening while most users know little about how search engines or AI systems operate.

Regardless of where people search, the implications extend far beyond algorithms. Personalization affects how teams collaborate, how data moves across departments, and how global organizations define success.

This article explores what personalization means today and how global brands can turn it into a competitive advantage.

From SERPs To AI Summaries

Search engines no longer return lists of blue links alone or People Also Ask (PAA). They now provide summarized information in AI Overviews and AI Mode, currently for informational queries.

Google often surfaces AI summaries first and URLs second, while continuously testing different layouts for mobile and desktop, as shown below.

Screenshot from search for [what is a nepo baby], Google, December 2025

Google’s Search Labs experiments, including features such as Preferred Sources, show how layouts and summaries change based on context, trust signals, and behavioral patterns.

Large language models (LLMs) add another layer. They adjust responses based on user context, intent, and sometimes whether the user has a free or paid account. Because users rarely get exactly what they need on the first attempt, they re-prompt the AI, creating iterative conversations where each instruction or prompt influences the next.

What prompts users to click through to a source or research it on search engines, whether it is curiosity, uncertainty, boredom, a call-to-action, or the model stating it does not know, is still unclear. Understanding this behavior will soon be as important as traditional click-through rate (CTR) analysis.

For global brands, the challenge is not simply keeping up with technology. It’s maintaining a consistent brand voice and value exchange across channels and markets when every user sees a different interpretation of the brand. Trust is now as important as visibility.

This landscape increases the importance of market research, segmentation, cultural insights, and competitive analysis. It also raises concerns about echo chambers, search inequality, and the barriers brands face when entering new markets or reaching new audiences.

Meanwhile, the long tail continues to shift to platforms like TikTok, where discovery works very differently from traditional search. And as enthusiasm for AI cools, many professionals believe we have entered the “trough of disillusionment” stage described by Jackie Fenn’s technology adoption lifecycle.

What Personalization Means Today

In marketing, personalization refers to tailoring content, offers, and experiences based on available data.

In search, it describes how search engines customize results and SERP features for individual users using signals such as:

Data patterns.
Inferred interests.
Location.
Search behavior.
Device type.
Language.
AI-driven memory (which is discussed below).

The goal of search engines is to provide relevant results and keep users engaged, especially as people now search across multiple channels and AI platforms. As a result of this, two people searching the same query rarely see identical results. For example:

A cuisine enthusiast searching for [apples] may see food-related content.
A tech-oriented user may see Apple product news.

SERP features can also vary across markets and profiles. People Also Ask (PAA) questions and filters may differ by region, language, or click behavior, and may not appear at all. For example, the query “vote of no confidence” displays different filters and different top results in Spain and the UK, and PAA does not appear in the UK version.

AI platforms push this further with session-based memory. Platforms like AI Mode, Gemini, ChatGPT, and Copilot handle context in a way that makes users feel there are real conversations, with each prompt influencing the next. In some cases, results from earlier responses may also be surfaced.

A human-in-the-loop (HITL) approach is essential to evaluate, monitor, and correct outputs before using them.

How Personalization Technically Works

Personalization operates across several layers. Understanding these helps marketers see where influence is possible.

1. SERP Features And Layout

Google and Bing adapt their layouts based on history, device type, user engagement, and market signals. Featured Snippets, PAA modules, videos, forums, or Top Stories may appear or disappear depending on behavior and intent.

2. AI Overviews, AI Mode, And Bing Copilot

AI platforms can:

Summarize content from multiple URLs.
Adapt tone and depth based on user behavior.
Personalize follow-up suggestions.
Integrate patterns learnt within the session or even previous sessions.

Visibility now includes being referenced in AI summaries. Current patterns show this depends on:

Clear site and URL structure.
Factual accuracy.
Strong entity signals.
Online credibility.
Fresh, easily interpreted content.

3. Structured Data And Entity Consistency

When algorithms understand a brand, they can personalize results more accurately. Schema markup helps avoid entity drift, where regional websites are mistaken for separate brands.

Bing uses Microsoft Graph to connect brand data with the Microsoft ecosystem, extending the influence of structured data.

4. Context Windows And AI Memory

LLMs simulate “memory” using context windows, which is the amount of information they can consider at once. This is measured in tokens, which represent words or parts of words. It is what makes conversations feel continuous.

This has some important implications:

Semantic consistency matters.
Tone should be unified across markets.
Messaging needs to be coherent across content formats.

Once an AI system associates a brand with a specific theme, that context can persist for a while, although it is unclear how long for. This is probably why LLMs favor fresh content as a way to reinforce authority.

5. Recommenders

In ecommerce and content-heavy sites, recommenders show personalized suggestions based on behavior. This reduces friction and increases time on site.

Benefits Of Personalization

When personalization works, users and brands can benefit from:

Reduced user friction.
Increased user satisfaction.
Improved conversion rates.
Stronger engagement.
Higher CTR.

This can positively influence the customer lifetime value. However, these benefits rely on consistent and trustworthy experiences across channels.

Potential Drawbacks

Alongside the benefits, personalization brings some challenges that marketers need to be aware of. These are not reasons to avoid personalization, but important considerations when planning global strategies. Consider:

Filter bubbles reduce exposure to diverse viewpoints and competing brands.
Privacy concerns increase as platforms rely on more behavioral and demographic data.
Reduced result diversity makes it harder for new or smaller brands to appear.
Global templates lose effectiveness when markets expect local nuance.

This means that brands using the same template or unified content across markets for globalization lose even more effectiveness in markets, as cultural nuance, context, or different user motivations are expected. Furthermore, purchase journeys vary across markets. Hence, the effectiveness of hyper-personalization.

It is probably more important than ever that brands spend time researching and planning to gain or maintain visibility in global markets, as well as strengthening their brand perception.

Managing Personalization Across Teams And Channels

At the moment, LLMs tend to favor strong, clearly structured brands and websites. If a brand is not well understood online, it is less likely to be referenced in AI summaries.

Successful digital and SEO projects rely on strong internal processes. When teams work in isolation, inconsistencies appear in data, content, and technical implementation, which then surface as inconsistencies in personalized search.

Common issues include:

Weak global alignment.
Translations that miss local relevance.
Conflicting schema markup.
Local pages ranking for the wrong intent.
Important local keywords being ignored.

Below is a framework to help organizations manage personalization across markets and channels.

1. Shared Objectives And Understanding Across Teams

Many search or marketing challenges can be prevented by building a shared understanding across teams of:

Business and project goals.
Issues across markets.
Search developments across markets.
Audience segmentation.
Integrated insights across all channels.
Data flows that connect global and local teams.
AI developments.

2. Strengthen The Technical Elements Of Your Website

Reinforce the technical elements of your website so that it is easy for search engines and LLMs to understand your brand across markets to avoid entity drift:

Website structure.
Schema markup on the appropriate sections.
Strong on-page structure.
Strong internal linking.
Appropriate hreflang.

3. Optimize For Content Clusters And User Intent, Not Keywords

Structure is everything. Organizing content into clusters helps users and search engines understand the website clearly, which supports personalization.

4. Use First-Party Data To Personalize On-Site Experiences

Internal search and logged-in user experiences are important to understand your users and build user journeys based on behavior. This helps with content relevance and stronger intent signals.

First-party data can support:

Personalized product recommendations.
Dynamic filters.
Auto-suggestions based on browsing behavior.

5. Maintain Cross-Channel Consistency

A coherent experience supports stronger personalization and prevents fragmented journeys, and search is only one personalized environment. Tone, structure, messaging, and data should remain consistent across:

Social platforms.
Email.
Mobile apps.
Websites and on-site search.

Clear and consistent USPs should be visible everywhere.

6. Strengthen Your Brand Perception

With so much online competition, brands whose work is being referenced positively across the internet. It is the old PR: Focus on your strengths and publish well-researched work, with stats that are useful to your target users.

Conclusion: Turning Personalization Into An Advantage

Conway’s Law matters more than ever. The idea that organizations design systems that mirror their own communication structures is highly visible in search today. If teams operate in silos, those silos often show up in fragmented content, inconsistent signals, and mixed user experiences. Personalization then amplifies these gaps even further by not being cited on AI platforms or the wrong information being spread.

Understanding how personalization works and how it shapes visibility, trust, and user behavior helps brands deliver experiences that feel coherent rather than confusing.

Success is no longer just about optimizing for Google. It is about understanding how people search, how AI interprets and summarizes content, how brands are referenced across the web, and how teams collaborate across channels to present a unified message.

Where every search result is unique, the brands that succeed will be the ones that coordinate, connect, and communicate clearly, both internally and across global markets, to help strengthen the perception of their brand.

More Resources:

Featured Image: Master1305/Shutterstock

Sponsored

Keep it together with Contentful

The go-to DXP for AI, Contentful manages everything from GEO to personalization, removing the roadblocks of modern marketing

Learn more