How We Verify Claims — Giyunu Gepiwa

The Problem with Most SEO Content

Someone writes a blog post. Another person reads it and writes their own version. A third person cites the second as a source. Within a few cycles, a speculation has become "established fact" — even though no one ever traced it back to Google's actual behavior.

This is how most SEO mythology propagates. Entity-based search is particularly vulnerable to this because the concepts are genuinely complex, the primary sources are technical, and there's a large market for simplified explanations that sacrifice accuracy for readability.

We take a different approach. Every factual claim on this portal must be traceable to a primary source before it appears in editorial content.

Sources We Accept

Google Patent Filings

USPTO and international patent applications filed by Google LLC or Alphabet Inc.

Peer-Reviewed Research

Papers published in academic venues, including Google Research and affiliated institutions.

Official Google Documentation

Google Search Central documentation, developer guides, and official blog posts.

Legal Proceedings

Court documents and regulatory filings where Google's systems are described under oath.

Our Process

From Raw Source to Published Article

Source Identification

We identify a primary source: a patent, a research paper, official documentation, or a legal filing. The source must come directly from Google, Alphabet, or an affiliated institution. Secondary sources are not used as the basis for factual claims.

Technical Reading

The source is read in full, not skimmed. For patents, this means reading the claims section, the detailed description, and the abstract. For research papers, this means reading the methodology, not just the conclusion.

Claim Extraction

Specific, bounded claims are extracted from the source. Vague implications are not treated as factual claims. If a patent describes a mechanism, we describe that mechanism — we don't extrapolate to behaviors the patent doesn't address.

Plain-English Translation

Technical content is translated into readable prose without losing accuracy. Where simplification requires omission of nuance, we note the simplification explicitly. Readers should always know when they're getting a simplified version.

Citation Embedding

Every factual claim is linked to its source in the published article. Patent numbers are cited in full. Research papers are referenced with their publication venue and year. Readers can verify every claim independently.

Patent Research

Reading Google Patents Responsibly

Google files hundreds of patents each year. Most are filed speculatively — they describe systems that may or may not be deployed. A patent grant does not mean a feature is live in search.

This distinction matters enormously. When we cite a patent, we describe what the patent claims, not what Google's search results necessarily do. We flag when a patent describes a system that appears to align with observable search behavior, and we flag when it remains speculative.

The value of patent research is not in revealing secret algorithms. It's in understanding the conceptual vocabulary Google uses internally — the categories, the scoring methods, the disambiguation approaches. That vocabulary helps you think about search the way Google thinks about it.

Researcher examining printed Google patent documents about entity recognition algorithms

Editorial Limits

What You Won't Find Here

Being clear about our limits is part of being trustworthy. These are things we deliberately exclude from this portal.

Optimization Services

This portal does not sell entity optimization, schema implementation, or any related service. Content is purely educational. There is no commercial interest in recommending any particular approach.

Fabricated Statistics

You will not find percentages or data points on this portal unless they come from a cited, verifiable source. "Studies show that X% of businesses..." without a source citation is not something we publish.

Guaranteed Outcomes

Search behavior is probabilistic and context-dependent. We don't claim that any particular action will produce any particular result. Understanding entity signals is different from controlling search outcomes.

Secondhand Interpretation Chains

We don't write articles that cite other SEO blogs as sources. If a claim can't be traced to a primary source, it doesn't appear as a factual claim. It may appear as an open question or an area of uncertainty.