More content can hurt SEO as AI search rewards authority

Google launched AI Overviews to all U.S. users in May 2024. Carolyn Shelby’s argument is blunt: as AI-driven search systems synthesize answers instead of counting documents, indiscriminate publishing can dilute authority rather than build it. That shift is forcing SEO agencies to replace volume targets with pruning, consolidation, intent mapping, and tighter quality thresholds.

Why the old scale playbook is breaking

For years, the logic behind SEO growth was straightforward: more pages created more keyword coverage, more long-tail reach, and more chances to rank. That model fit a document-centric search engine, where a large library could outperform a smaller one even when many pages were only average. AI-driven retrieval changes that math because systems evaluate relationships, semantics, and entity signals across a site, not just the existence of another URL.

Google’s automated ranking systems prioritize helpful, reliable, people-first content, not pages created mainly to manipulate rankings. Core updates are broad changes Google makes several times a year, and meaningful sitewide quality improvements can take months to show up in search results. For agencies, the penalty for weak content is no longer just low performance on one page. It can drag down how an entire site is read.

Google’s product changes made the shift harder to ignore

By October 28, 2024, Google said AI Overviews were expanding to more than 100 countries and territories and reaching more than 1 billion monthly users. By May 2025, Google said AI Overviews were available in more than 200 countries and territories and more than 40 languages.

In 2025, Google said people were asking longer questions and exploring deeper topics through AI Overviews, which it framed as a new discovery opportunity. At the same time, Google’s March 2024 spam and low-quality-content updates were expected to reduce low-quality, unoriginal content in search results by 40%, a figure later revised to 45%. The system is expanding answer-first experiences while also tightening the screws on unoriginal material.

Google Search Central’s crawl-budget guidance is specifically aimed at large, frequently updated sites and treats crawl optimization as a best practice. Site bloat is also an efficiency problem, because wasted URLs can consume crawling and refresh attention that should be reserved for the pages with the strongest authority and commercial value.

What replaces volume publishing

The practical answer is not “publish nothing.” It is to treat content as an architecture problem instead of a production calendar problem. The strongest agencies are moving toward content pruning, consolidation, intent mapping, and explicit quality thresholds that decide whether a page deserves to exist, be merged, or be rebuilt.

A useful way to think about the change is this:

• Pruning removes thin, stale, or overlapping pages that no longer earn their keep.
• Consolidation merges duplicate articles and similar service pages into stronger, more complete assets.
• Intent mapping aligns each page to a search need, so informational, comparative, and transactional queries do not compete with one another.
• Quality thresholds force a page to clear a standard for originality, depth, entity coverage, and internal linking before it is published.

At Ahrefs and Semrush, content pruning and duplicate-content cleanup remain practical SEO tactics, which fits the current environment. Cleanup alone does not define a stronger information architecture. Those tools are especially useful when a site has years of content debt and needs to identify pages that are cannibalizing one another or adding little topical value.

Moz’s 2026 coverage emphasizes entity clusters, trust, and AI visibility. That framing is better for agencies that need to show clients how topics connect across a domain, not just which posts attracted clicks. It still depends on disciplined execution, clean internal linking, and a willingness to remove content that looks active but functions as clutter.

How agency retainers and teams have to change

A retainer built around monthly post quotas is hard to defend when Google is rewarding helpfulness, compactness, and sitewide coherence. The better-fit offer is an ongoing program of audits, content inventories, pruning sprints, consolidation work, and topic architecture planning.

Team structure changes too. Agencies that once relied on high-volume content production need more editors, technical SEOs, and information architects who can judge overlap, intent gaps, and entity relationships. Writers still matter, but they are less valuable as output engines and more valuable as specialists who can deepen a cluster, strengthen a page’s semantic clarity, and support the pages that actually carry authority.

Reporting has to evolve with that shift. Raw publish counts and word totals are weak proof in an AI search environment. Stronger reporting tracks how many pages were removed or merged, which topic clusters gained coverage, whether important URLs are being crawled and refreshed efficiently, and how visibility changes as AI Overviews expand. Agencies are no longer selling traffic from a pile of posts. They are selling authority signals that survive answer-first search. That aligns with Search Engine Land’s 2026 view of search, where AI handles discovery, decisioning, and transactions.