Clicky

The Definitive Guide to SEO Audits post-Google Leak

EEAT is mapped to most likely attributes. See EEAT Decoded.

I guess you could call this my opus. This is also my last post on the hobo seo blog, bringing almost a 20-year blogging run to a close. That’s almost as long as Andy Dufresne took. Reason? I’ll get into trouble if I keep decoding the black box in public. The purpose of the decoding was always to get to this point. It was always about white hat seo auditing. It’s not about abusing Google’s systems. It’s about working within them.

And it’s been a busy week or so uncovering the erstwhile too hidden Goldmine scoring system, the Firefly system and Craps.

I also had a go at debunking LSI keywords, creating the definitive guides post-leak for on-page seo, title tag seo, meta tag seo, image SEO, internal links and local SEO. I also published EEAT Decoded and mapped algorithm updates to Google leak attributes. This work builds upon the seminal work of Mike King, Rand Fishkin and my own deep investigation over the last 3 months, with my own logical inferences.

This framework translates the leaked Google Content Warehouse attributes and systems into a highly detailed and actionable SEO audit framework for any website.

This SEO framework absorbs leak and exploit analysis, Google Search Essentials, the Google Search Quality Evaluation guidelines, the DOJ antitrust trial revelations, and statements from official Google spokespeople (Googlers).

Each task is directly mapped to the underlying technical signals Google uses to evaluate content, allowing for an SEO strategy that is aligned with Google’s core engineering principles.

So this could be the first actual evidence-based SEO audit framework. An SEO audit based on the Content Warehouse leaked attributes.

It’s why I made it.

Section 1: Trust & Safety Audit (T)

Of these aspects, trust is most important. The others contribute to trust, but content doesn’t necessarily have to demonstrate all of them. For example, some content might be helpful based on the experience it demonstrates, while other content might be helpful because of the expertise it shares.” Google 2025

Trust

This section audits the foundational signals that determine if a site is technically sound, safe, and provides a satisfying user experience.

A failure in this section can trigger punitive demotions or prevent a site from being considered for competitive rankings in the first place.

This entire section contributes to a site’s foundational, query-independent (Q*) score. – which sworn testimony in the DOJ Vs Google Antitrust Trial revealed to be a site’s overall measure of “trustworthiness“.

Hobo SEO Dashboard.
How Google deals with Spam. Screenshots from Hobo SEO Dashboard.

Before you embark on any SEO project, I recommend you first review a foundational ranking factor: the trustworthiness of the site you are auditing.

See 2.5.2: Finding Who is Responsible for the Website and Who Created the Content on the Page.

Then proceed:

Task 1: Audit User Satisfaction & The Navboost Feedback Loop

 “The ranking itself is affected by the click data. If we discover that, for a particular query, hypothetically, 80 percent of people click on Result No. 2 and only 10 percent click on Result No. 1, after a while we figure out, well, probably Result 2 is the one people want. So we’ll switch it.” Google 2012

Avoid users pogo sticking

Why it Matters: User behaviour is the ultimate validation of trust and relevance. The Navboost system re-ranks results based on a 13-month window of click data. This data, processed by the CRAPS (Click and Results Prediction System), analyses signals like GoodClicks, BadClicks, lastLongestClicks, and unsquashedClicks (vetted, genuine interactions).

Raw popularity is also measured via signals like chromeInTotal (total site views from Chrome), which feeds into the top-level Popularity (P)* signal.

Google’s Goldmine system (that I discovered through my own journalistic-type research – as amplified by Barry at Search Engine Roundtable) creates a direct feedback loop: a low-quality title can lead to poor user clicks, which feeds negative data into Navboost via the goldmineNavboostFactor, which can then demote your page’s core ranking.

How to Audit:

  • Analyse SERP Performance: In Google Search Console, identify pages with a low Click-Through Rate (CTR) relative to their ranking position, as this can contribute to a serpDemotion. Use a tool like Hobo SEO Dashboard Winners and Losers reports to analyse pages in bulk and quickly identify pages demoted in SERPS. 
  • Analyse On-Page Behaviour: Use Google Analytics (or a little tool I like, Clicky) to find pages with high bounce rates or low dwell times. These are indicators of user dissatisfaction that can lead to BadClicks and a navDemotion.
  • Map User Journeys: For key landing pages, audit the internal links. Do they provide logical pathways to help a user fully satisfy their query within your site, or do they lead to dead ends? Screaming Frog Spider is a good tool for deep auditing of internal links.
  • Audit page quality: Use the Prompt “Rate My Page Quality Using The Hobo SEO Method“.

Task 2: Audit for Mobile Experience & On-Page Clutter

Google uses the mobile version of a site’s content, crawled with the smartphone agent, for indexing and ranking. This is called mobile-first indexing… Don’t let ads harm your mobile page ranking. Follow the Better Ads Standard when displaying ads on mobile devices.” Google 2025

mobile SEO
Testing your page experience on a mobile browser is your no1 priority.

Why it Matters: A poor user experience is a direct, punitive signal that negatively impacts your site’s Q* score.

The SmartphonePerDocData module contains a boolean violatesMobileInterstitialPolicy for intrusive pop-ups and a scaled integer adsDensityInterstitialViolationStrength (0-1000) for the severity of ad clutter.

These contribute to a site-level clutterScore that can be “smeared” (isSmearedSignal) across the entire domain, suppressing its ranking potential.

A page must also pass the isSmartphoneOptimized check and avoid legacy negative signals like a high maximumFlashRatio. A similar violatesDesktopInterstitialPolicy flag confirms these penalties are not exclusive to mobile.

How to Audit:

  • Eliminate Intrusive Interstitials: Audit your site on mobile and desktop devices to find and remove any pop-ups or interstitials that cover the main content in a way that contravenes Coalition for Better Ads and therefore Google’s guidelines.
  • Reduce Ad Density: Ensure that ads do not dominate the “above-the-fold” content on mobile, forcing users to scroll excessively to find what they are looking for.
  • Confirm Mobile-Friendliness: Manually review the site on a mobile device to ensure pages are correctly optimised for smartphones.

Task 3: Audit Indexing Tiers & Foundational Accessibility

Why it Matters: A document can be disqualified before ranking even begins.

The SegIndexer system places documents into quality-based tiers like “Base” (high-quality), “Zeppelins,” or “Landfills” (low-quality), determined by its scaledSelectionTierRank.

Being relegated to the Landfills severely limits a page’s ability to ever rank. This is the first and most critical gate in the ranking pipeline.

How to Audit:

  • Ensure Indexability: Verify pages are not blocked by robots.txt or noindex tags. This is the absolute basic requirement to even be considered for indexing.
  • Perform a Site-Wide Quality Review: A high volume of low-quality, thin, or spammy content risks having significant portions of the site classified in low-quality indexing tiers. Prune and improve this content.

Task 4: Audit for Content Quality Debt (Panda System)

The Google Algorithm Update Correction Hypothesis (Shaun Anderson, Hobo)
The Google Algorithm Update Correction Hypothesis (Shaun Anderson, Hobo)

Why it Matters: The legacy of the Google Panda updates exists as a persistent, site-wide pandaDemotion signal (and its variants babyPandaDemotion & babyPandaV2Demotion) that acts as “algorithmic debt” for domains with prevalent low-quality, thin, or duplicate content.

This is technically identified using signals like shingleInfo to detect duplicate text, a general lowQuality score, and a vlq (“Very Low Quality”) model score.

How to Audit:

  • Perform a Content Inventory: Crawl the entire site with Screaming Frog SEO Spider to create an inventory of all content.
  • Identify Low-Quality Pages: Use analytics and crawler data to identify pages with low traffic, low word count, or minimal user engagement. Hobo SEO Dashboard automatically works with Screaming Frog and identifies these types of pages (called Dead Pages).
  • Remediate or Remove: Have a plan to either significantly improve, consolidate, or strategically remove (with proper redirects) the identified low-quality content to lift the potential pandaDemotion.

Task 5: Audit for Spam & Deception Signals

E-E-A-T
E-E-A-T

Why it Matters: A trustworthy site must be free of manipulative tactics.

Google employs a suite of classifiers, including a scamness score, GibberishScore, urlPoisoningData, spamrank (for linking to bad neighbourhoods), KeywordStuffingScore, and a general DocLevelSpamScore.

How to Audit:

Task 6: Audit for Initial Trust & Sandbox Effect

Why it Matters: New URLs present a “cold start” problem for click-based systems like Navboost.

The tofu (Trust on First Use) signal is a predictive quality score for new URLs based on intrinsic signals like site-wide quality and technical health.

The hostAge attribute is used to “sandbox fresh spam,” providing the technical basis for the initial period of limited visibility for new domains.

How to Audit:

  • Optimise for “Trust on First Use”: Ensure new content is regularly published on a technically sound site (HTTPS, good Core Web Vitals) with a history of quality and a clean outbound link profile to earn a high initial tofu score.
  • Be Patient with New Domains: Understand that the hostAge signal means new domains will likely face an initial evaluation period before they can achieve full ranking potential.

Task 7: Audit Foundational Technical Trust & Canonicalisation

Why it Matters: A technically sound foundation is a direct trust signal. The system includes an explicit demotion for a badSslCertificate.

Clean canonicalisation is also critical to avoid duplicate content (ContentChecksum96) and ensure all signals are consolidated into a single, clean CompositeDoc (the master record for a URL).

The system also checks for technical errors with flags like isErrorPage and isWebErrorMobileContent.

How to Audit:

  • Verify SSL Certificate: Ensure your SSL certificate is valid, properly installed, and covers all domain variations
  • Audit Redirects & Canonicals: Use a site crawler like Screaming Frog to check for redirect chains and ensure rel=”canonical” tags are implemented correctly to consolidate authority and prevent signal fragmentation.

Task 8: Audit for Local Entity Verification (for local businesses)

Why it Matters: For local entities, real-world trust is quantified.

The LocalWWWInfo Module uses signals like brickAndMortarStrengthScore (a score for a business’s physical prominence), cluster, and wrapptorItem for entity resolution to verify a business’s identity.

This is related to the concept of geotopicality, where a business can inherit authority from its physical location.

How to Audit:

  • Ensure NAP Consistency: Audit your business listings across the web to ensure your Name, Address, and Phone number are consistent.
  • Optimise Google Business Profile: Keep your business hours, address, and other information on your Google Business Profile meticulously accurate and up-to-date.

A page can demonstrate expertise, but if it is untrustworthy, it will be assigned the lowest possible quality rating because, as the September 11 2025, guidelines state, “Trust is the most important member of the E-E-A-T family because untrustworthy pages have low E-E-A-T no matter how Experienced, Expert, or Authoritative they may seem”.

Section 2: Experience Audit (E)

“Originality is something we’re going to be focusing on this year. That’s going to be important,” Gary Iilyes, Google 2025

This section audits for signals that demonstrate first-hand, real-world experience with a topic. This is Google’s algorithmic defence against generic, low-effort, and unoriginal content and a key input to the Q* system.

Task 9: Audit Content Effort, Originality & Linguistic Quality

Why it Matters: Google’s systems act as automated quality editors. contentEffort is an “LLM-based effort estimation” that measures the difficulty of replication.

This is complemented by the OriginalContentScore. Furthermore, systems like BlockBERT (goldmineBlockbertFactor) and SnippetBrain perform deep semantic analysis, evaluating content for linguistic quality and coherence, not just keyword matching.

How to Audit:

  • Assess Irreplaceability: For your key content, ask: “Does this page contain proprietary data, original research, or unique insights that cannot be easily replicated by AI or a competitor?” Use the Prompt “Rate My Page Quality Using The Hobo SEO Method“.
  • Prioritise Human-Readable Content: Ensure content is well-written, grammatically correct, and flows logically to score well with semantic analysis systems like BlockBERT.

Task 10: Audit for Clear Authorship Signals

Why it Matters: Google explicitly tracks creators.

  • The system uses a boolean isAuthor flag to confirm if an author was identified and an author attribute to store an encoded identifier for that person. This is further supported by E-E-A-T signals like authorObfuscatedGaiaStr, which links content to a specific author entity.

How to Audit:

  • Implement Author Bylines: Ensure every article has a visible author byline.
  • Create Detailed Author Pages: Link each byline to a comprehensive author biography page that details their credentials, experience, and publications.

Task 11: Audit Content Freshness & Algorithmic Momentum

Why it Matters: The system rewards meaningful content maintenance.

The lastSignificantUpdate attribute records the timestamp of the last major revision.

A sophisticated view of freshness is built from bylineDate, syntacticDate, and semanticDate, and processed by systems like FreshnessTwiddler. Core signals are stored as a VersionedFloatSignal, meaning Google tracks a page’s quality trajectory over time, creating “algorithmic momentum”.

How to Audit:

  • Focus on Substantive Changes: When updating content, focus on adding new information or updating data to ensure the revision is flagged as “significant”.
  • Identify QDF Topics: Prioritise significant update efforts on topics that are time-sensitive and likely to benefit from a freshness boost.

Task 12: Audit for Visual Prominence and Original Multimedia

Why it Matters: Visuals serve as evidence of first-hand experience. Image SEO is incredibly important.

The docImages array contains an original_media_score that assesses whether images are unique or generic stock photos. Additionally, visual prominence is a measured signal via the avgTermWeight attribute, which quantifies the “average weighted font size of a term in the doc body.”

How to Audit:

  • Reduce Stock Photo Reliance: Replace generic stock photos with custom graphics and original photography to increase the original_media_score.
  • Ensure Visual Hierarchy: Check that your main <h1> is the most visually prominent title on the page and that headings are enclosed in the correct semantic <h*> tags.

Task 13: Audit for Specialised Content Quality

Why it Matters: Google uses dedicated systems for specific content types.

For product review sites, an entire suite of signals including productReviewPUhqPage (“Ultra High Quality Page”), productReviewPPromotePage, productReviewPDemoteSite, productReviewPReviewPage, productReviewPDemotePage, and productReviewPPromoteSite is used to reward deep, first-hand testing.

For sites with forums or comments, the ugcScore and ugcDiscussionEffortScore are used to distinguish valuable discussions from low-quality spam.

How to Audit:

  • For Review Sites: Ensure product reviews demonstrate deep, first-hand use with original photos and unique data.
  • For UGC Sites: Foster and moderate on-page communities to encourage substantive, high-effort conversations.

Task 14: Audit Page Speed & Core Web Vitals

Core Web Vitals Components
Core Web Vitals Components

Why it Matters: Poor CWV scores lead directly to negative behavioural signals (like badClicks) that Navboost penalises.

The IndexingMobileVoltVoltPerDocData module explicitly uses mobileCwv and desktopCwv for “ranking changes.” The IndexingMobileVoltCoreWebVitals module stores the field data metrics lcp, inp, cls, and the legacy fid. Google also collects performance data like time-to-first-byte-per-doc directly via Chrome.

How to Audit:

  • Monitor Field Data: Use the Core Web Vitals report in Google Search Console as your primary source of truth, as it uses real-world user data (Field Data). The goal is to pass the “Good” threshold for at least 75% of users.
  • Debug with Lab Data: Use PageSpeed Insights and Lighthouse to run Lab tests to diagnose and fix the root causes of poor LCP, INP, and CLS scores.

Section 3: Expertise Audit (E)

Expertise

This section audits for signals that measure the depth of knowledge and topical specialisation of a website. A strong performance here is critical for both the Q* (Authority) and T* (Relevance) systems.

Task 15: Audit for “Signal Coherence”

Why it Matters: The Goldmine engine, which powers the T* relevance system, thrives on consistency.

It scores title candidates based on their alignment with other page elements, using factors like goldmineBodyFactor (relevance to content) and goldmineUrlMatchFactor (alignment with URL), and giving special weight to the <h1> via goldmineHeaderIsH1. To make your intended title the undeniable winner, all signals must be in harmony.

This is the “B” (Body) of the “ABC” signals.

How to Audit:

  • Perform a Cross-Element Review: For your most important pages, verify that the HTML <title>, the main <h1>, the URL slug, and the introductory paragraph all send a consistent, harmonised message about the page’s core topic.
  • Structure for Machines: Implement comprehensive Schema.org markup to populate the richsnippet field and make your content eligible for rich results.

Task 16: Audit for Semantic Comprehensiveness (QBST & Entities)

Why it Matters: For any given query, the QBST (Query-Based Salient Terms) system has “memorised” a list of concepts it expects to find on a comprehensive page.

This is supported by the EntityAnnotations module, which attaches specific Knowledge Graph entities (webrefEntities) to your content. The asteroidBeltIntents system also suggests a highly granular model for classifying document intent.

How to Audit:

  • Reverse-Engineer the SERPs: For a target query, analyse the top 5-10 ranking pages. Identify the common subheadings, entities, and concepts they all share. This is your blueprint for the salient terms QBST expects.
  • Define Entities Clearly: Use structured data (Schema.org) and clear, unambiguous language to help Google’s EntityAnnotations process correctly identify the key entities your content discusses.

Task 17: Audit Topical Focus & Coherence

Topical Authority Strategy
Topical Authority Strategy

Why it Matters: Google algorithmically rewards specialisation.

The site2vecEmbeddingEncoded attribute creates a numerical representation of your site’s theme, which powers the siteFocusScore (quantifying niche dedication) and siteRadius (measuring deviation from the core topic).

A high siteFocusScore and low siteRadius for key pages are measurable signs of topical authority.

How to Audit:

  • Map Your Content Clusters: Analyse your site architecture to ensure content is logically grouped into tight, thematic clusters (hub-and-spoke model).
  • Identify and Prune Outliers: Identify content that is thematically distant from your core expertise. Consider improving it to be more relevant, or pruning it if it detracts from your siteFocusScore.
  • Learn about Topical authority: See my article: Topical Authority: Topicality & Popularity Post-Google Leak.

Task 18: Audit Internal Prominence & Linking.

Pagerank described via a heating system metaphor
Site structure is described via a heating system metaphor.

Why it Matters: A site demonstrates its own expertise through its internal linking structure. The onsiteProminence attribute measures internal link equity by simulating user traffic flow, signalling which pages the site itself considers most important and authoritative on a topic.

How to Audit:

  • Review Internal Linking Strategy: Ensure your most important “cornerstone” pages have the most internal links pointing to them from relevant, contextual anchors.
  • Use a “Hub and Spoke” Model: Structure content in topical hubs where a main pillar page links out to more specific sub-pages, and those sub-pages link back to the pillar.

Task 19: Audit for YMYL (Your Money or Your Life) Content

Why it Matters: Google applies higher standards for high-stakes topics. The chard classifier is likely an initial predictor used to identify the nature of content, including whether it falls into the YMYL category. If so, it triggers a more rigorous evaluation using specific classifiers like ymylHealthScore and ymylNewsScore.

How to Audit:

  • Identify YMYL Content: Audit your site to identify any pages that offer financial, legal, or medical advice.
  • Apply Highest Standards: For this content, ensure information is factually perfect, all sources are meticulously cited, and the author’s expert credentials are prominently displayed.

Section 4: Authoritativeness Audit (A)

This section audits for signals of reputation and standing within an industry. This is the pinnacle of the Q* (Quality) score.

Task 20: Audit Site-Level Authority & Algorithmic Momentum.

Why it Matters: Google uses a persistent, site-wide quality score.

A primary input is the siteAuthority score. This score can be increased by an authorityPromotion boost or actively penalised by an unauthoritativeScore.

This system is built on the Normalised Site Rank (NSR) framework, where predictedDefaultNsr serves as the baseline quality score. Because this is a VersionedFloatSignal, Google tracks your site’s quality trajectory over time, creating “algorithmic momentum.”

How to Audit:

  • Adopt a Long-Term View: Recognise that siteAuthority is a long-term, composite metric. Consistent, long-term investment in quality is algorithmically rewarded.
  • Review Information Architecture: Ensure your site has a clean, logical structure (e.g., Home > Categories > Products) that is easy for users and crawlers to navigate.

Task 21: Audit Link-Based Authority & Anchor Text as a Title Candidate

Google Pagerank Updates Explained

Why it Matters: Links remain foundational.

The modern evolution of PageRank is PageRank-NearestSeeds, and Homepage PageRank (homepagePagerankNs) functions as a trust anchor for the entire domain.

The legacy toolbarPagerank and PageRankPerDocData attributes confirm the long history of this system. The numOffdomainAnchors attribute measures link volume.

Critically, the Goldmine system uses anchor text from both internal (sourceOnsiteAnchor) and external (sourceOffdomainAnchor) links as candidates for a page’s title in the SERP. This is the “A” (Anchors) of the “ABC” signals.

How to Audit:

  • Shift Focus to Linking Pages: When building or analysing links, focus on the linking page’s potential for user traffic. The mantra is to “Earn Clicks on Pages That Link”.Audit Anchor Text for SERP Fitness: Review your internal and external anchor text. Is it a concise, compelling, and accurate description that you would be happy to see as your page’s title in the search results?

Task 22: Audit Anchor Text Relevance & Profile

Why it Matters: Anchor text is a “very important signal” for Google’s foundational relevance scoring system (T*). The system applies a punitive anchorMismatchDemotion when the anchor text does not align with the topic of the target page.

How to Audit:

  • Ensure Topical Relevance: Audit internal links to ensure anchor text is descriptive and strongly aligned with the content of the destination page.
  • Assume First Link Priority: When multiple links on a page point to the same destination, assume that only the first link’s anchor text is considered for relevance scoring and plan accordingly, but do not rely on it or be bound by it.

Task 23: Audit for Domain-Level Keyword Penalties

Why it Matters: The system includes specific penalties for domain-level manipulation. The exactMatchDomainDemotion is a direct penalty for low-quality sites that use an exact-match keyword as their domain name, neutralising this tactic for sites that lack genuine quality.

How to Audit:

  • Assess Site Quality: If your site is on an EMD, you must be hyper-vigilant about quality.
  • Overcompensate with E-E-A-T: Ensure your content quality, user experience, and authority signals are exceptionally high to counteract any potential negative bias from this demotion factor.

Task 24: Audit Author & Brand Entity Signals

Why it Matters: The system directly measures brand authority. The authorObfuscatedGaiaStr attribute links content to specific author entities. The queriesForWhichOfficial signal stores the specific queries for which a page is considered the definitive “official” result, a powerful brand signal associated with the Vince update.

How to Audit:

  • Build Author Entities: Encourage authors to build their own public profiles and expertise on relevant platforms.
  • Target “Official Source” Status: For your core brand and product terms, ensure your content is so comprehensive and authoritative that it becomes the undisputed official source. This will attract links.

Task 25: Audit for Brand Scale & Scope

Why it matters: While the specific attributes isLargeChain and siteSiblings were not detailed in the provided sources, the concept of classifying businesses based on their operational model is a key part of authority evaluation. Google sets different authority expectations for a large national brand versus a small local business.

How to Audit:

  • Use Structured Data: Implement Organization schema markup to clearly define your brand, its official name, logo, and any parent/subsidiary relationships.
  • Ensure Consistency: Make sure your brand is represented consistently across all your web properties and third-party profiles.

Section 5: Image SEO Audit

This section audits for signals related to how Google understands, evaluates, and ranks visual content, based on the ImageData schema.

Task 26: Audit Image Provenance & Indexing

Why it Matters: Google’s architecture is obsessed with identifying the original source of an image. The contentFirstCrawlTime timestamp is used to identify the first time an image’s content was seen anywhere on the web, giving preference to the original. Not all images are indexed; a system called “Amarna” (corpusSelectionInfo) acts as a quality gate. Even among duplicates, a hierarchy exists, measured by rankInNeardupCluster.

How to Audit:

  • Prioritise Original Imagery: Use unique, original images whenever possible to establish your site as the source of truth.
  • Ensure High Technical Quality: Use high-resolution, well-compressed images to ensure they pass the initial “Amarna” quality gate and are selected for indexing (isIndexedByImagesearch).

Task 27: Audit In-Image Semantic Content.

Why it Matters: Google achieves semantic mastery by analysing the pixels themselves. Multiple OCR systems (ocrGoodoc, ocrTaser) read and index text within images. Object recognition (imageRegions) and entity linking (multibangKgEntities) connect objects in your image directly to Google’s Knowledge Graph.

How to Audit:

  • Use Information-Rich Graphics: Create charts, diagrams, and infographics with clear, legible text that can be extracted via OCR.
  • Use Entity-Dense Photography: Produce photographs that depict multiple, contextually relevant entities in a clear relationship with one another (e.g., a photo of a specific landmark, not a generic building).

Task 28: Audit Image Quality & Aesthetics.

Why it Matters: Google algorithmically scores an image’s beauty and technical merit. A Neural Image Assessment (NIMA) model provides a technical quality score (nimavq) and an aesthetic appeal score (nimaAva). A high whiteBackgroundScore is used as a proxy for professional product photography and commercial trust.

How to Audit:

  • Invest in Professional Visuals: Use well-lit, sharp, and well-composed images to achieve high algorithmic quality scores.
  • For E-commerce: Use clean, white backgrounds for primary product shots to signal professionalism and trustworthiness.

Task 29: Audit for User Engagement & Clickbait Signals.

Why it Matters: Real-world user behaviour is the ultimate arbiter of an image’s quality. The system tracks granular metrics like h2c (hovers-to-clicks) and h2i (hovers-to-impressions). Critically, a clickMagnetScore actively penalises images that attract clicks from irrelevant “bad queries,” directly fighting visual clickbait.

How to Audit:

  • Optimise Thumbnails for Relevant Clicks: A/B test different image crops and compositions to identify which versions generate the highest click-through rates from relevant queries.
  • Avoid Misleading Visuals: Ensure your images are an honest representation of the page’s content to avoid attracting “bad queries” and triggering a high clickMagnetScore.

Task 30: Audit for Commercial & Licensing Data.

Why it Matters: Commerce is a native function of image search. The shoppingProductInformation attribute is the backend data store for rich product results. The imageLicenseInfo field directly powers the “Licensable” badge in search results, driven by metadata embedded in the image file (embeddedMetadata) or page schema.

How to Audit:

  • For E-commerce: Ensure flawless implementation of Product and MerchantListing structured data on all product pages.
  • For Content Licensors: Establish a workflow for embedding comprehensive IPTC metadata (creator, copyright, licensing URLs) into every image file before upload.

Task 31: Audit Image Safety & Policy Compliance.

Why it Matters: Google uses a sophisticated, multi-layered system for content moderation. An initial computer vision analysis (brainPornScores) is combined with contextual signals (like the queries an image ranks for) to produce a finalPornScore. Specialised detectors like hateLogoDetection proactively identify harmful content.

How to Audit:

  • Ensure Brand Safety: Regularly audit the queries your images are ranking for in Google Search Console to ensure they are not being associated with inappropriate or unsafe topics.
  • Maintain Policy Compliance: Ensure all visual content adheres to Google’s SafeSearch policies to avoid being flagged or removed.

Section 5: Navigating Google Algorithm Updates

By following the advice in this definitive guide to SEO audits post Google leak, you will avoid Google algorithm update heartache. Check out my post on mapping leaked Google attributes to algorithm updates.

Google algorithm updates mapped to potential leaked signal attributes.
Google algorithm updates mapped to potential leaked signal attributes.

 

To complete your understanding of SEO audits in the post-leak world, check out my article on on-page SEO.

Hobo
Privacy Overview

This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.