Key Takeaways
- Sharp ranking swings rather than gradual declines, especially affecting informational sites and generic SaaS blogs built around summarising existing SERP content.
- Articles that repeat existing SERP explanations without adding new insights are losing their AI Overview visibility.
- Sites exceeding 200ms are experiencing noticeable mobile ranking losses, particularly in competitive niches.
- Anonymous “Admin” authorship is collapsing in YMYL sectors. Finance and health pages now strongly favour identifiable experts with verifiable LinkedIn or industry credentials.
- Superficial fixes are no longer enough. Recovery requires pruning thin AI-scaled content and consolidating topical authority.
With the March 2026 Core Update, Google is no longer rewarding perfectly optimised articles that simply summarise existing information. Instead, the algorithm is prioritising Information Gain, meaning content must provide something new, such as original data, first-hand experience, or local regulatory insights that generic AI-generated articles cannot replicate.
As an AI-driven SEO agency based in Malaysia, we have been closely monitoring the rollout since early March. So if you’re wondering why your rankings or impressions dropped, you come to the right place (because we were wondering too).
In this article, we will share what we have observed so far, what appears to be driving ranking volatility, and how SEO specialists can adapt their strategy to remain visible in search.
Let’s break it down.
Table of Contents
The “SDR” Comparison: Recovery vs Traditional SEO
Strategy Component | Traditional SEO (Pre-2026) | March 2026 Recovery Strategy |
Content Focus | Keyword density and article length | Information Gain and original data |
Technical KPI | LCP and CLS optimisation | INP below 150ms |
EEAT Signal | Generic author bio | Verifiable expert with LinkedIn or industry citations |
AIO Strategy | Schema markup | Atomic answers and proprietary visuals |
Link Building | Domain Authority (DA) focus | Citations from AIO-featured sources |
How the March 2026 Core Update Changed SEO Strategy
For years, SEO strategies focused on technical optimisation, backlinks, and long-form content.
Those factors still matter, but the March update shows us that Google now wants sources rather than summaries.
In practical terms, this means:
- Articles must contribute new insights, not repeat existing ones.
- Pages must demonstrate real experience and authority.
- Content must be structured so AI systems can extract clear, concise answers.
Sites that fail to meet these criteria are increasingly losing visibility in both traditional rankings and AI-generated results.
“You need something unique, something fresh, something that no one has written before that Google did not capture.” – Riff Chen, regional manager of Rankpage
Why Your Traffic Dropped (The “Consensus Trap”)
Many sites are losing traffic because their content repeats what already exists on the SERP without adding new information.
Since the rollout began on March 1st, we have audited more than 30 websites across different industries, including our own sites and clients.
One pattern appears repeatedly:
- Sites experiencing a 30–50% traffic drop are not necessarily poorly optimised.
- Their content simply duplicates the same explanations already present across multiple websites.
But Google’s updated Synthesis Engine now identifies clusters of articles that say essentially the same thing.
For example:
If ten blogs explain that DOT certification is a safety standard for helmets and motorcycles, Google’s AI can summarise those explanations into a single answer box. Once that happens, users no longer need to click through to any of those ten sites.
However, we have observed something interesting.
When a website introduces a local correction or additional context, visibility often increases rather than declines.
For example:
- A global article might explain DOT certification requirements.
- A Malaysian-focused article might clarify that DOT certification is not recognised by Malaysia’s JPJ regulatory framework. (which is true)
That small piece of local insight transforms the article from consensus content into Information Gain.
In the past the SEO Skyscraper technique was very popular among writers, take the existing top ranking blogs, write better and done, you rank higher. Whatever they have, our blog also needs it.
Now, it’s not the case.
“Instead of building taller skyscrapers than your competitors, you need to build a different building altogether, does that make sense?” – Nelson Hon, Senior Editor at Rankpage.
Why Lower-Ranked Pages Are Appearing in AI Overviews
AI Overviews now evaluate standalone answer quality separately from traditional rankings.
Pages ranked between positions 20–50 can still appear in AI summaries if they provide clearer answers, stronger structure, or unique insights that higher-ranking pages lack.
Previously, most AI Overview citations came from the Top 10 organic results. That made sense because ranking strength and answer quality were strongly correlated.
Since the March 2026 update, Google appears to run a separate extraction layer that evaluates if a page contains a better standalone answer, regardless of its ranking position.
In other words:
- Traditional rankings decide which pages appear in the SERP.
- The AI extraction system decides which pages provide the best answer units.
When those two signals diverge, a page ranked in position 35 can still become the citation source.
What Google Seems to Be Evaluating Now
From the pages we analysed, the AI Overview extractor appears to prioritise four signals.
Signal | What It Looks For |
Atomic Answers | Short, direct answers placed immediately under headings |
Information Gain | New insights not repeated across other SERP pages |
Structured Content | Tables, lists, and clear section hierarchy |
Authority Signals | Named authors, citations, and credible references |
Pages that satisfy these criteria can be selected for AI summaries even if their traditional ranking signals are weaker.
This explains why some well-optimised articles lost their citations even though they continued ranking on page one.
The “Extraction Layer” Hypothesis
Based on our testing, we believe Google is now running something like a two-stage retrieval system.
Stage 1: Traditional Ranking
The search engine identifies relevant documents using standard signals such as:
- Backlinks
- Keyword relevance
- Domain authority
- Page experience
This produces the familiar Top 10 search results.
Stage 2: AI Extraction
A second system scans a wider pool of pages and asks a different question.
Instead of “Which page ranks best?” it asks:
“Which page contains the best standalone answer?”
That answer unit is then inserted into the AI Overview.
Because this system searches beyond the Top 10, pages ranked between 20 and 50 can still be selected.
Why Some High-Ranking Pages Lost Citations
In many cases, the Top 10 pages simply failed the Information Gain test.
When multiple articles repeat the same explanation, the AI system only needs one representative source. The rest become redundant.
For example: Ten articles explaining “What is a payment gateway” will likely produce the same definition.
However, if one article includes:
- Local regulatory insight about Bank Negara Malaysia
- FPX integration requirements
- PDPA requirements
That page becomes significantly more useful for a Malaysian audience.
The AI Overview will then favour that page even if it ranks lower in traditional results.
The Hidden Ranking Signal: Entity Credibility
One pattern we noticed during our audits is that Google appears to be placing greater emphasis on verifiable entity credibility.
In the past, it was common for blogs to publish articles under generic author names such as:
- Admin
- Editorial Team
- Content Team
And we are guilty of that too.
However, many of these pages are now losing visibility, particularly in sensitive niches such as finance, health, and legal advice.
Google appears to favour content that is connected to real professionals with verifiable expertise.
Signals that seem to matter include:
- LinkedIn author profiles
- Industry certifications
- Company knowledge graph associations
- Mentions on credible third-party sites
This has always been part of Google’s long-standing EEAT framework, but the March update appears to enforce it more aggressively. Presumably, because of many, many AI personas.
For Malaysian businesses, this means content should ideally be associated with:
- Industry professionals
- Company founders
- Subject matter experts
Rather than anonymous editorial profiles.
What This Means for SEO Strategy
The implication is important because ranking high is no longer enough to guarantee AI visibility.
Content must also be designed for extraction.
From our audits, the pages most likely to appear in AI Overviews tend to include:
- Short direct answers under each heading
- Tables
- Concise bullet lists
- Real-world examples or internal data
- Credible institutional references
These elements make it easier for AI systems to extract clear answer segments.
An Example
Imagine two articles targeting the same query.
Article A
- 2,500 words
- optimised keywords
- strong backlinks
- long narrative paragraphs
Article B
- 1,200 words
- Clear answer blocks
- Structured tables
- A short case study
- References to local authorities
Article A may still rank higher in organic search.
But Article B may be selected for the AI Overview because it contains better extractable answers.
What We Are Doing Differently Now
After observing these patterns, Rankpage had started adjusting how we structure our content.
Instead of focusing purely on ranking signals, we now design articles around atomic information blocks that can be easily extracted.
For example:
- Direct answers under headings
- Clearly labelled comparison tables
- Short evidence sections
- “Field notes” explaining real observations
This approach increases the likelihood that an article becomes a source for AI-generated summaries rather than just another ranking page.
Our Action Plan for Recovery
If your analytics dashboard suddenly turned red in early March, we recommend starting with a structured Information Gain audit.
Based on our observations, the following three steps have produced the most consistent recovery signals.
Prune the “AI Scale” Pages
Pages that contain generic AI-generated information without unique insights can weaken your entire domain’s authority.
Look for pages that show:
- Moderate impressions
- Low engagement
- No original data
- Explanations identical to competing articles
Ask yourself a simple question.
Could an AI model generate this page in ten seconds using publicly available information?
If the answer is yes, the page is likely contributing little Information Gain.
Our recommended actions:
- Consolidate overlapping articles into one comprehensive master guide
- Remove thin pages that dilute topical authority
- Redirect valuable backlinks to stronger pages
This process improves crawl efficiency and reinforces the credibility of your remaining content.
Inject “Proof of Work”
Google increasingly values real-world experience signals.
For every major article, we recommend adding a Field Notes section that demonstrates first-hand observations.
This can include statements such as:
“In our audit of Malaysian SME websites in 2025, we found that pages with structured answer blocks increased AI Overview citation visibility by approximately 28 percent.”
Supporting elements may include:
- Screenshots from analytics dashboards
- Internal testing data
- Workflow diagrams
- Case study metrics
These additions serve two purposes.
First, they strengthen Experience signals within EEAT, which is always good.
Second, they provide information that AI models cannot easily reproduce, increasing the article’s Information Gain.
- Anchor to Local Entities
Local regulatory context has become an important authority signal.
We have observed ranking improvements when content references credible Malaysian institutions such as:
- Malaysian Communications and Multimedia Commission (MCMC)
- Inland Revenue Board (LHDN)
- Road Transport Department (JPJ)
- Department of Statistics Malaysia (DOSM)
Linking advice to these institutions demonstrates that your content is grounded in verifiable regulations and real-world context.
In several cases, we observed 15–20% visibility improvements after adding authoritative references and clarifying regulatory implications for Malaysian businesses.
Conclusion: Is SEO Dead? No, so Don’t Freak Out
The March 2026 Core Update represents a shift away from content that simply aggregates existing information and while we are guilty of that too, we learned from our mistakes.
Websites that rely on repeating the same explanations found across dozens of other blogs will increasingly struggle to maintain visibility .
That means:
- Publishing original insights
- Documenting lived-in experiences
- Providing local context that automated content cannot replicate.
But that is easier said than done, and we know the struggles of it.
At Rankpage, we have been actively adjusting our AI SEO strategies to align with these changes, focusing on Information Gain, extractable content structures, and origins; insights.
- Key takeaways
- Direct Answer
- Quested based H2s
- Now adding infographics
We’ve been doing that since 2025 and slowly, we notice other agencies are doing that too.
As search continues to shift toward AI-powered discovery, we believe the future belongs to agencies that adapt quickly.
That is exactly what we are preparing for.
Frequently Asked Questions About March Core Update
What Is the March 2026 Core Update?
The March 2026 Core Update is a broad algorithm adjustment prioritising Information Gain, stronger technical performance such as INP optimisation, and verifiable author expertise.
How Long Does It Take to Recover From the March Update?
Based on our observations, recovery typically takes four to twelve weeks after implementing significant Information Gain improvements, as Google reassesses overall site trust.
Should I Delete AI-Generated Content?
Not necessarily. Content should only be removed if it provides no unique value. In many cases, adding original data, expert commentary, and practical examples can significantly improve its performance.
Why Is Information Gain Important After the March 2026 Update?
Information Gain refers to content that adds something new rather than repeating existing explanations. Google now prioritises pages that provide original insights, local context, or first-hand experience instead of content that simply summarises other websites.
Does Page Speed Still Matter After the Update?
Yes. Technical performance remains important, but the focus has shifted toward Interaction to Next Paint (INP). Sites with slower interaction responsiveness, especially above 200ms, are increasingly seeing mobile ranking declines and weaker visibility in AI-generated search results.
Can Smaller Websites Still Rank After the March 2026 Update?
Yes. Smaller websites can still gain visibility if they publish unique insights and structured answers. In some cases, niche sites are appearing in AI Overviews because their content provides clearer explanations or local expertise that larger sites fail to include.
