Let's be blunt: the era of throwing cheap AI articles at the wall to see what sticks is dead. If you are still relying on raw, automated volume to win search traffic, you are running a playbook that is already obsolete.
Here is the hard truth. Recent search data reveals that 60% of Google queries now end without a single click to any content. Let that sink in for a second. More than half of the search traffic you think you are targeting never actually reaches your website. When generative tools make publishing practically free, the old thesis of “more is better” collapses. It doesn't just fail to earn traffic; it actively dilutes your brand's authority, adding noise to a channel that is already deafening.
Search Engine Journal—which has grown from a simple personal blog back in 2003 into an industry institution drawing over 710,000 monthly visitors under the direction of parent Alpha Brand Media—recently hosted a webinar detailing this shift. During the session, Gabriel Dillon, Contentful’s GTM Lead for Personalization, and John Graham, Contentful's Principal Solution Strategist, walked through why AI-assisted copy drifts toward generic output and how to fix it. Dillon’s core message was simple: volume is no longer a viable pipeline strategy. The only content that moves the needle today is content built for a specific reader, held strictly accountable to real business outcomes, and verified by real-time data.
— Adam Riemer
Source
- Approximately 60% of Google searches now result in zero clicks.
- Gabriel Dillon from Contentful argues that content volume alone is no longer an effective digital strategy.
- Content must be tied to specific business outcomes and human intent to succeed.
Source: https://www.searchenginejournal.com/about/
- Search Engine Journal (SEJ) was founded as a personal blog in 2003.
- SEJ is owned by Alpha Brand Media and averages over 710,000 monthly visitors.
The "Ultimate Yes Man" Problem
Why does copy generated by AI assistants feel so remarkably uniform? Because your assistant is the ultimate "yes man."
When we prompt these systems, we feed them our own assumptions and unverified biases. They map our inputs against the average of what already exists on the web. The result? A self-reinforcing echo chamber of mediocrity. Instead of gaining fresh analysis, we end up mirroring our direct competitors, validating what we already wanted to hear, and ultimately leaving the reader empty-handed. It validates what we already believe but provides no real value.
Dillon notes that the missing ingredient isn't a better prompt. It's "taste." Not in some abstract, design-school way, but as concrete discernment and risk-taking. It is the human willingness to state a point of view that no generic model would volunteer on its own. AI can synthesize context at lightning speed, but the author must supply the conviction.
"Our biases as we write content using the robots ends up eating the content that we produce," Dillon explained. "We end up in this cycle of creating content that we think is good but doesn’t actually do what we think it does."
If you want to survive this shift, you have to stop publishing blindly. As detailed in SEO’s publishing golden rule is dead — here’s why more content now hurts your visibility, flooding your domain with unoriginal pages will actually trigger algorithmic penalties and destroy your search visibility.
— Adam Riemer
Source
- AI writing tools mirror user biases and competitive drafts, leading to generic content convergence.
- Human taste, discernment, and proprietary market insight are required to break the loop.
Four Questions to Fix Your Content Accountability
If you are not holding your content accountable—a rising complaint from clients who feel agencies are just delivering words without pipeline impact—you are wasting resources. Let’s be honest: tracking rankings without tying them to business growth is just vanity. Before you greenlight another draft, run it through Gabriel Dillon's four-part filter.
These four questions should serve as your content gatekeepers:
- Does this copy produce our expected outcomes?
- Who is this content actually for?
- How do we identify those specific people?
- How does this insight scale?
If you cannot answer these with hard numbers, you do not have a strategy. You have a text document. Experimentation should not be an occasional project you run once a quarter; it needs to be a core operational system. Mapping content variations to actual business milestones builds a loop that keeps you from publishing stuff just to hit an arbitrary deadline.
"If we don’t have data that proves that our content is good, then we can’t really think about the way to scale it out or make it more effective," Dillon warns.
— Adam Riemer
Source
- Dillon runs a four-question accountability filter on all marketing copy before publication.
- Continuous experimentation and measurement prevent teams from producing empty volume.
Personalization Without the Headache
Why do B2B personalization programs collapse? Because teams build massive, over-engineered systems that are impossible to maintain. They spend six months configuring enterprise suites only to find out they have no data to run them.
You do not need a machine learning team to make personalization work. Focus on the data you are already collecting. Dillon structures this into three straightforward tiers:
- Tier 1: New vs. Returning Visitors. This is low-hanging fruit. A first-time visitor needs different copy and distinct CTAs compared to someone who has visited your site four times. Acknowledge where they are in their buyer journey.
- Tier 2/3: Ad Campaign and Loyalty Hubs. These are gold mines. Instead of searching for complex external signals, look at the UTM parameters and customer details your ad platforms and loyalty systems are already capturing.
The goal is clear differentiation, not micro-segmentation perfection. Dillon demonstrated in the Contentful walkthrough that building these setups is straightforward if you do not overcomplicate the underlying technology.
— Adam Riemer
Source
- Overcomplicating personalization stacks causes enterprise marketing programs to stall.
- The three tiers of personalization signals starts with visitor status (new vs. returning) and ad campaign tags.
- Simple content platforms can deliver personalization without custom engineering.
Competing in the Zero-Click Search Layer
Forget trying to defeat Google's AI detectors. The search engine is too smart, and trying to outrun their systems is a race you will not win. Instead, focus on the real problem: AI search summaries are absorbing the user clicks that used to go directly to your pages.
To win, you have to compete directly within the AI answer layer. Generative Engine Optimization (GEO) and AI Experience Optimization (AEO) are the new standards. The guidelines for high-quality, high-integrity content that ranks in traditional search are the exact same parameters that AI engines rely on to compile their summaries. If your technical architecture is solid, your site will feed the LLMs. But if it is broken, you do not exist to the machine. As explored in Why SEO Remains the Essential Infrastructure for AI Search Success, your crawling foundation sits at the core of all AI discovery.
If your executive team is pushing for high-volume, automated AI content, show them the performance data. Make the case that publishing fewer, higher-quality pieces yields better conversion rates.
— Adam Riemer
Source
- Organic traffic declines as AI answer layers intercept search queries directly.
- Competing in the AI answer layer requires aligning traditional technical SEO foundation with generative optimization.
Tackling AI Bias and Leadership Demands
The webinar's Q&A segment touched on the core challenges search marketers are facing right now.
For instance, user concerns about Google's spam updates removing AI content. Dillon's advice is clear: do not focus on evading detection. The true problem is traffic loss from zero-click layouts.
Then there is the issue of bias in AI output. Dillon points out that bias enters in two distinct ways: the training data the models were built on, and the prompts we write. When we prompt the system to write what we want to hear, we guarantee a mediocre result. To fix this, you must run quality checks and build your strategic perspective before generating a single word.
Lastly, what do you do when your leadership team demands infinite volume? You show them the math. Compare the conversion rate of a few highly focused, original pieces against the flatline results of a massive AI-slop campaign. Show them that in a world where search is shifting, more pages do not equal more customers.
— Adam Riemer
Source
- Evading AI content detection is a temporary battle; the real focus must be on zero-click search.
- AI bias is introduced via training sets and prompt constraints.
- Content teams must demonstrate with data that quality-focused content programs outperform high-volume AI campaigns.
- Standard utility pages like pricing or service lists do not require a strong character voice, but must still serve clear user intents.
