Google’s June 2025 Core Update: Subtle Shifts Driven by New Tech

The rollout of Google’s June 2025 Core Update has concluded. While some experts described it as a major algorithmic shift, the overall impact appeared less dramatic compared to previous updates. This suggests Google may have made behind-the-scenes adjustments rather than visible overhauls to the search experience.

Behind the Curtains: New Tech Likely in Play

Two developments in Google’s technological stack might help explain what’s changed. While many speculated the update was tied solely to MUVERA — a newer retrieval algorithm — there’s more beneath the surface. Alongside MUVERA, Google recently introduced a Graph Foundation Model (GFM), both of which are likely to influence ranking outcomes.

MUVERA: Refining How Results Are Retrieved

MUVERA stands for Multi-Vector via Fixed Dimensional Encodings, and it aims to streamline how search results are retrieved. Traditional multi-vector search models often cast a wide net, pulling in a large number of pages. MUVERA, in contrast, narrows down the selection to fewer, more relevant documents — improving efficiency and accuracy.

This system has the potential to enhance precision without compromising performance. It reduces computational load while still surfacing high-quality content, meaning less noise in results and better matches for users’ queries.

GFM: Smarter Understanding of Relationships

The Graph Foundation Model, introduced in July 2025, represents another leap forward. Unlike traditional graph neural networks, which are limited by the datasets they’re trained on, GFM can generalize across unfamiliar structures. This adaptability allows it to analyze connections between content, users, and links more holistically.

Already tested internally — particularly in identifying spam in ads — GFM has shown it can outperform legacy models by interpreting the relational data in new, more insightful ways. This means Google could now better understand content trustworthiness and relevance, even in unfamiliar contexts.

Why This Matters for Rankings

Although Google didn’t confirm that these models were deployed as part of the June update, their timing strongly suggests some influence. MUVERA could be helping Google better filter results, while GFM may allow it to spot trustworthy content more effectively. Combined, they support Google’s goal of providing results that users genuinely find helpful.

The update didn’t cause massive fluctuations across the board. Some websites previously impacted by earlier changes — like the Helpful Content Update — saw rankings recover, while others slipped further. But the overall reaction from the SEO community was muted, suggesting a more measured impact.

The Bigger Picture

Rather than a disruptive shake-up, this update seems more like a refinement — tuning the system to better detect quality, relevance, and authority. With advanced retrieval methods and smarter relational models, Google continues moving toward a search experience that prioritizes usefulness over quantity.