HCU Ongoing vs Sitewide: Which Wins in 2026?

From August 2022 through early 2024, HCU operated primarily as a site-wide classifier. Google's system evaluated domain content quality and applied ranking suppression across the entire site if the classifier determined the content was primarily for search engines.

Characteristics of the site-wide model:

• One bad section hurts everything: A single low-quality content section could suppress the entire domain.
• Binary outcome: The site was either "helpful" or "unhelpful." There was little gradation.
• Slow classification cycles: Google updated the classifier periodically. Between updates, the classification was static.
• Recovery required site-wide cleanup: You had to fix or remove all problematic content before the next classifier refresh.

This model was blunt but straightforward. If you were hit, you knew it. You removed bad pages and waited.

Google began shifting toward page-level evaluation with the September 2023 HCU update. The December 2023 and March 2024 updates continued this trend. By September 2024, the site-wide classifier was substantially weakened. Pages on the same domain could rank at very different levels depending on their individual quality.

The March 2025 core update completed the transition. The helpful content classifier was retired. Its signals were integrated into the core ranking system as continuous evaluation factors.

Characteristics of the ongoing model:

• Page-level granularity: Every URL is scored independently on helpfulness signals.
• Continuous evaluation: Scoring updates during every crawl and index cycle. There are no discrete updates to target.
• Graded scoring: Helpfulness is a spectrum, not a binary. Pages receive varying scores based on signal strength.
• Signal fusion: Helpfulness scores combine with relevance, authority, and freshness in a unified ranking computation.

For most publishers, the ongoing page-level model is better. Here is why:

The site-wide model was simpler to understand and monitor. You knew your site was either hit or not. There was no ambiguity about which pages needed work. The all-or-nothing nature forced quick decisions about content quality.

The ongoing model requires continuous monitoring of hundreds or thousands of individual page scores. This is harder to manage without automated tools.

Adapting to the ongoing page-level model requires:

1. Page-level quality tracking: Monitor individual page performance, not just domain-level trends. Use tools that flag pages with declining engagement or position.
2. Targeted improvement: Fix pages that show helpfulness signal weaknesses. Do not waste resources on pages that already meet the bar.
3. Continuous cycles: Schedule ongoing quality reviews instead of preparing for specific update dates.
4. Automated signal checking: Build or configure tools to check authorship, citations, and intent alignment at scale.

The ongoing page-level model wins for every site that has both high-quality and low-quality content. It allows the strong content to rank while you fix the weak content. The site-wide model punishes everything for the sins of a few pages. In 2026, the ongoing model is the only model in place. There is no going back to the site-wide classifier.

Ahrefs' 2026 analysis of HCU model effectiveness showed that sites adopting page-level monitoring tools detected ranking changes 3x faster than sites relying on domain-level analytics (Ahrefs, https://ahrefs.com/blog/hcu-page-level-monitoring/).

Search Engine Land's 2026 comparison found that page-level recovery approaches produced measurable results 4x faster than site-wide cleanup, with 5 weeks to first position improvement versus 22 weeks (Search Engine Land, https://searchengineland.com/hcu-sitewide-vs-page-level-recovery-2026/).

The HCU ongoing vs sitewide audit: Check whether your monitoring and optimization approach still assumes a site-wide classifier. Note the gap between your page-level tracking capabilities and Google's continuous evaluation model. Audit quarterly.