Lies & More Lies

Google

Google Has Been Lying About Their Ranking Algorithm - Is Anyone Surprised? 

Some of Google’s internal API documentation was accidentally made public on GitHub from March 27 - May 7, 2024, and the leaked algorithm conflicts with past statements from Google regarding what factors affect search ranking. Erfan Azimi, an SEO expert and CEO of EA Eagle Digital, discovered the 2500 page leak and shared his findings with Rand Fishkin, a seasoned SEO veteran. The documentation was corroborated by former Googlers as consistent with Google’s standards of practice, naming conventions, and other internal processes. 

After weeks of painstaking analysis and verification, Fishkin published his findings to inform people of the reality of Google’s criteria, dispelling myths that were propagated by the company itself and uncovering truths that seem intentionally buried. 

• Click-Centric User Signals: Despite Google's public denials, the documents suggest Google uses clickstream data extensively, including clicks, click-through rates (CTR), long vs. short clicks, and user engagement signals.

• Subdomain and Sandbox Usage: The documentation contradicts Google's denial of treating subdomains separately and using a sandbox for newer websites.

• Domain Age: Contrary to Google's public statements, the leak suggests domain age is collected and heavily considered.

• Whitelists for Sensitive Topics: Covid-19 and Elections: Google employed whitelists during the Covid-19 pandemic and democratic elections to control which sites appeared prominently in search results.

• Link Quality Tiers: Google categorizes links into quality tiers (low, medium, high) based on click data, influencing their impact on rankings.

• User Intent Scoring: NavBoost scores queries based on user engagement, influencing the appearance of videos or images for specific queries.

• Post-Query Engagement: Engagement metrics from follow-up searches impact rankings of associated keywords and websites.

• Classic Ranking Factors: PageRank and anchors are less influential, but Page Titles remain significant.

• E-E-A-T Considerations: Expertise, Authoritativeness, and Trustworthiness (E-E-A-T) may not be as directly influential in rankings as previously thought, outside of author authority. 

• Site Focus Measurements: Focused content that relates to a specific niche seems to benefit rankings when clear topic authority is demonstrated.

The Response

How should these findings affect your SEO strategies as a business leader?

• Brand Importance: Building a notable brand outside of Google search is crucial for ranking success. Small and medium businesses may struggle in SEO unless they establish significant credibility and user demand elsewhere. 

• User Intent: User engagement and navigational demand are powerful ranking factors, potentially overriding traditional SEO signals. Have click worthy headlines that deliver on their promise. 

• Establish Authority: Focused, quality content with measurable value for the user and a clear demonstration of topic authority will serve you better than large troves of AI-generated, keyword-stuffed blogs.

Rabbit R1

A Hare-Brained Gimmick

The Rabbit R1, like all AI hardware of the moment, paints an incredible vision for the future but fails to provide any value in the now. It’s meant to be an assistant of sorts that eliminates the need for users to interface with an app by performing tasks based solely on voice commands.  

Theoretically, users could say “Order me an Uber home” or “Play my Workout Playlist” and the R1, which is logged in to apps like Spotify, DoorDash, and Uber, would perform those tasks for you. In reality, the gadget comes back with an error or buffering message, performs the task incorrectly, or just fails to do anything at all. 

Though the Rabbit R1 is priced fairly reasonably at $199, its capabilities are so underwhelming those 2 bills are far better spent elsewhere, like on an acai bowl or a gallon of milk in this economy. There are exactly 0 things the R1 can do better than a smartphone, outside of offering a satisfying button to click. 

AI gadgets are an exciting possibility for the future, but at this moment in time they are superfluous gimmicks compared to the still wildly convenient, widely utilized smartphone.

AI Overview

Source: Trust Me Bro

Google’s AI Overview feature has been spitting out some outrageous answers, suggesting readers eat rocks and put glue on pizza. Google reports that users have been more satisfied with their search results since the launch of the AI Overview feature, and attributes the majority of the “odd responses” in question to faked screenshots – in classic Google fashion. Despite Google’s attempts at shifting blame, we know that generative AI has proven to be unreliable and slightly silly at times, so users should be cautious taking any AI output at face value. 

As the “old ways” of Web2.0 slowly begin to disappear, it’s important that we as business leaders approach these new paradigms with caution and an appropriate amount of skepticism, while still keeping an open mind to the innovations that can change how we do things for the better. 

How do you feel about the rise of AI - good, bad, and ugly? Let us know in the Reality Bytes Linkedin group, linked below!