Google Core Update : New Research Explains the Losses

A comparative analysis of 113 websites hit by the 2024 Google Core Updates finds technical SEO, not AI content or backlinks, is the primary driver of traffic losses. Here is what the data shows.

This post builds on our earlier analysis of technical SEO issues correlating with lost rankings after the last Google Core Update. The current update was still rolling out at the time of writing, with websites bouncing up and down in the SERPs. So we ran methodological research on 113 affected sites to understand what was really driving the changes.

---

Research Methodology

We applied a Comparative Analysis methodology, a standard approach in empirical research combining three techniques:

Sample Selection Criteria

We analyzed websites belonging to our clients and websites posted in SEO communities where owners complained about traffic losses. The key selection rule: the owner had to blame the loss on any cause except technical SEO. This prevented sampling bias toward sites already aware of our hypothesis.

Additional criteria: losses had to occur on sites that were already ranking, and the traffic drop had to be at least 50%. That threshold sounds high, but it reflects what is actually happening. We also contacted colleagues who contributed data under strict NDA agreements. The total sample reached 113 websites.

---

Google Core Update: First Insights

Before presenting the technical findings, we documented how site owners themselves explained their traffic losses. Figures exceed 113 because most sites cited multiple causes.

AI content (42 sites): Nearly one third blamed AI-generated content, even though Google publicly stated this was not a ranking signal. Given the cultural prominence of generative AI, this attribution is understandable but largely unsupported by our data.

Backlinks (37 sites): The second most cited cause. Owners questioned either the quality of their own links or whether competitors had acquired stronger ones. Neither explanation accounts for complete demotions or pages being pushed past position 100.

On-Page SEO (36 sites): Mostly overlaps with the AI content concern, focusing on content quality, header structure, and search intent alignment.

Ads (29 sites): Owners suspected excessive advertising. Notably, 21 of these sites used interstitial ads or intrusive popups, which is a documented user experience problem and merits further investigation.

User Experience (17 sites): We counted only those where the owner explicitly used the terms “poor user experience,” “poor UX,” or “UI/UX issues.”

CMS (8 sites): A small group questioned whether their choice of CMS or static code put them at a disadvantage relative to custom-built sites.

Unknown (83 sites): The largest group simply did not know. This is expected: you cannot identify a cause without a proper diagnostic.

Technology Stack of Affected Sites

We also recorded the technology used to build each affected site. WordPress powers over 40% of the web and roughly 70% if all CMSs are combined. Based on those market-share figures, we would expect approximately 80 WordPress sites and 33 non-CMS or custom-coded sites in a sample of 113. The actual distribution was strikingly different.

Metric	Value	Change	What it means
WordPress	12 sites	Expected ~80	Significantly underrepresented relative to 40%+ market share.
Custom Coded / JS	39 sites	~5–10	Massively overrepresented. 19 of these used ReactJS in CSR mode.
Wix	5 sites	Near expected
Shopify	3 sites	Near expected
Squarespace	2 sites	Near expected
Bubble.io	2 sites	Near expected
Undisclosed	43 sites	Likely JS or static HTML	Digital markers ruled out WordPress, Wix, Shopify, and Bubble.io.

Of the 39 custom-coded or JS-based sites, 19 used ReactJS, almost 50% of that group. Given React’s actual market share of roughly 4.2%, this represents approximately 320% overrepresentation in our affected sample. That figure has a logical explanation, covered in the conclusions below.

---

When Did the Traffic Losses Actually Happen?

Around 60% of site owners attributed their losses in part to the 2023 Helpful Content Update (HCU), describing a pattern along the lines of: “it started with the HCU, continued through the March Core Update, and my traffic finally disappeared with the August Core Update.”

We had access to Google Search Console and Google Analytics data for the majority of cases, either as guests or via screen captures. In a minority of cases we relied on owner-reported figures.

Case Examples: HCU, GCU, Both, and Neither

Case 1: Site Attributed to HCU

This site ran entirely on AI-generated content. The owner attributed the drop to that content, then tried removing it, writing posts manually, and purchasing backlinks. None of those interventions restored traffic.

A quick console inspection revealed a different story. The site had multiple JavaScript errors, including attempted Web3 injections, and roughly half its resources were failing to load. Core Web Vitals scores appeared near-perfect because a development agency had optimised for perceived speed by loading an empty shell first, then pulling in resources asynchronously. Google was crawling the shell, not the content.

The owner never suspected a technical cause because the speed metrics looked excellent. Content changes had no effect. The site never recovered its rankings, confirming that the real cause was technical, not content quality.

Case 2: Real Hit by Both HCU and GCU

This site was genuinely affected by the HCU, then by the March and August 2024 Google Core Updates. The webmaster applied fixes between each update, but each fix only held until the next update arrived. The root cause was ReactJS running in Client-Side Rendering (CSR) mode, compounded by a probable database issue that required further testing to confirm.

Case 3: Losses Misattributed to Updates

This site recorded its highest traffic during the HCU and GCU windows, then lost it one or two months later. Because ranking changes typically materialise during or immediately after an update, this lag rules the updates out as a direct cause.

Our working theory: the update temporarily boosted the site to test user behaviour. When behavioural signals showed no improvement, the site was demoted. Follow-up investigation confirmed why: the site had severe UX issues that prevented any user from completing the intended flow, resulting in zero conversions. The owner confirmed this diagnosis and began a full redesign.

The pattern repeated across many sites where we had access to analytics: traffic losses were incorrectly attributed to updates. The Core Web Vitals report for this site showed most metrics in red, including an 11-second Largest Contentful Paint.

---

Is the Google Core Update Technically Focused?

After analysing all 113 sites, our initial hypothesis holds: the statistical significance of the data confirms that the 2024 Google Core Updates were primarily focused on technical SEO and user experience, not content, backlinks, or authority.

On the sites hit only by the HCU (11 sites), all attempted some form of fix. Only 3 were subsequently hit by a GCU. Of the 9 hit by both HCU and GCU, 4 tried content-only fixes. All 4 failed. This strongly suggests that the GCU surfaced technical issues the content fix never addressed.

Across the 42 sites affected by GCU1, GCU2, or both, 35 attempted recovery. Of those:

• 18 saw short-term improvement, then were hit again by the next update.
• 9 achieved a clear, sustained improvement, reaching or exceeding pre-loss traffic.
• 8 were hit by the August GCU, so outcomes remained undetermined at time of writing.

What Was Actually Causing the Losses?

We audited all 113 sites individually, focusing exclusively on evidence that confirmed or refuted a technical SEO cause. The results:

We offered a technical fix to every site in the confirmed and likely categories, whether they were our own clients, community members, or colleague referrals. Three reasons: to thank contributors for sharing data, to test whether the fixes would improve traffic, and to demonstrate that technical corrections are fast to implement and produce near-immediate results. No more “wait six months”: technical fixes are boolean. They either work or they do not.

Of the 84 sites with confirmed or likely technical issues: 37 were already fixed, 28 were unknown (either no follow-up data or too early to measure), and 19 remained unfixed for various reasons.

---

Conclusions: What the Data Shows

Based on this sample of 113 websites, the March and August 2024 Google Core Updates were primarily targeting technical SEO and user experience, not content quality, backlink profiles, or domain authority.

Two technology findings stand out. First, WordPress sites were not specifically targeted despite holding the largest CMS market share. Second, ReactJS was heavily overrepresented in affected sites at roughly 320% above its expected frequency. The most probable explanation is architectural: ReactJS in Client-Side Rendering (CSR) mode delivers a minimal initial payload. Google crawls that payload and may not process the dynamically loaded content, or may penalise sites for large layout shifts caused by deferred loading. We recommend migrating to a Server-Side Rendering (SSR) approach when using React-based frameworks.

On AI-generated content: we found no data demonstrating any influence, positive or negative. Content improvements also failed to produce sustained recoveries, even for sites originally hit by the HCU. The most plausible interpretation is that Google identified additional technical issues beyond content, and content-only fixes never resolved the underlying problem. For a broader view of how content and technical factors interact, see our analysis of the 2026 core update research.

---

Final Thoughts

With so many pages now carrying near-identical information, partly because of AI generation at scale, content has become less differentiating as a ranking signal. The HCU was an early signal of this shift. The 2024 GCUs appear to be the logical continuation: when content parity is high, technical execution becomes the tiebreaker.

For SEOs and site owners, the practical implication is straightforward. Unlike content authority, which takes months to build, technical SEO issues are relatively fast to fix and their results are almost immediate. If your site lost traffic in 2024, a rigorous technical SEO audit is the highest-leverage starting point. You can also review our recovery guide for Google quality updates and our case study on three sites we recovered after HCU and GCU penalties.