This month, we bring you some of the latest developments from the world of search, including the highly anticipated roll-out of AI Overviews and the leak of Google’s Search API documentation.
Firstly, the leaked Google Search API document has been a major talking point within the SEO industry. This document details over 2,000 modules and 14,000 attributes, offering a glimpse into the extensive data Google collects. We’ll discuss what this leak means and how it should be approached.
In another significant update, Google has begun rolling out its AI Overviews to users in the UK. A comprehensive study by Onely, which analysed 25,000 e-commerce search queries, aims to shed light on the impact of this new feature. In this article, we’ll discuss the key findings of the study and shed light on how this may impact you.
Additionally, Google has clarified that the algorithmic actions targeting site reputation abuse have not yet been launched. This clarification comes amid widespread speculation that sites were already being penalised under this policy.
We’ll delve into these developments in detail below.
Allow our traffic light system to guide you to the articles that need your attention, so watch out for Red light updates as they’re major changes that will need you to take action, whereas amber updates may make you think and are definitely worth knowing but aren’t urgent. And finally, green light updates which are great for your SEO and site knowledge but are less significant than others
Keen to know more about any of these changes and what they mean for your SEO? Get in touch or visit our SEO agency page to find out how we can help.
In this post, we’ll explore:
- Leaked Google API documents cause a stir
- Google’s AI Overviews shake up ecommerce search visibility
- What does this mean for ecommerce brands?
- What can be done to try and feature in an AI Overview?
- Google confirms: No algorithmic actions for site reputation abuse yet
- The Helpful Content Update system is being retired
- An overview of the previous helpful content system
- How the helpful content system has changed in light of the core update
- Advice for SEOs and website owners
- May: A volatile month for SERPs
During May, there was an apparent leak of Google Search API documents, causing a stir within the SEO community.
A total of 2,569 modules and 14,014 attributes were mentioned in the leaked documentation. Initially shared by Erfan Azimi, these documents raised suspicions about Google’s transparency regarding its ranking signals. However, these concerns have since diminished, with many in the SEO community advising caution in interpreting the leak and its content.
The authenticity of the documents is still uncertain, though former Google employees have indicated they appear legitimate. The leaked document, titled “Document AI Warehouse Overview,” contains a vast amount of information about the search engine’s attributes and modules. While it lacks context on the provided information and the weighting of features, it reveals significant details about the data Google collects. Some SEOs hastily concluded that the document listed all of Google’s ranking signals, but this is inaccurate. The document includes some ranking factors, but others are not, so caution should be taken when interpreting the document.
In an article published by ipullrank, a careful analysis of the leaked document is provided. King contextualises the information contained to provide a more rounded summary of what the document really means.
Key takeaways from the document include:
- Site Authority: Despite Google’s denial of Domain Authority as a ranking factor, the document mentions a feature called “siteAuthority.” Its measurement and impact on rankings are unclear, but its existence and monitoring by Google are confirmed.
- Click Monitoring: Google tracks “badClicks” and “goodClicks.” The purpose and usage of this data are unknown, but considering the role of Navboost, clicks likely influence ranking algorithms.
- Host Age: Although Google has previously dismissed the idea of a Sandbox preventing new sites from ranking well, the document references “hostAge,” a concept used “to sandbox fresh spam during serving time.”
- Algorithmic Demotions: The document details several demotion types, including:
- Anchor Mismatch: Links that do not align with their target sites are demoted.
- SERP Demotion: Factors like user dissatisfaction or high bounce rates might lead to demotion.
- Location Demotions: “Global” and “super global” pages can be demoted, suggesting Google associates pages with specific locations for ranking purposes.
While intriguing, these insights highlight the nuanced nature of Google’s ranking systems, showing the need to carefully interpret and contextualise any information provided by the search engine.
A recent study conducted by Bartosz Góralewicz, founder of Onely, analysed 25,000 ecommerce search queries in an attempt to discover the impact of the roll-out of Google’s AI Overviews.
This study discovered that in the US:
- 16% of ecommerce queries now feature AI overviews in search results, accounting for 13% of the total search volume in this sector.
- 80% of the sources listed in these AI overviews do not rank organically for the original query. This implies a major shift away from traditional organic rankings.
Ranking in positions 1-3 only gives you an 8% chance of being shown in an AI Overview.
What does this mean for ecommerce brands?
According to international SEO consultant Aleyda Solis, Google is focusing on an “accelerated” approach, showcasing direct results instead of summarising ranked pages. This means traditional rankings (e.g., category pages, reviews, buying guides) are being bypassed.
This shift can be unsettling for established brands but opens new opportunities for visibility without targeting the most commercially valuable keywords.
What can be done to try and feature in an AI Overview?
In order to feature in an AI Overview, ecommerce retailers have been advised to optimise product data and detail pages to fit the AI overview format, which can offer visibility opportunities without competing for top organic rankings.
To assess AI Overview traffic impact, Solis has created a Google AI Overviews (AIO) Traffic Risk Impact Assessment Sheet. This sheet can be used to assess and analyse the potential traffic impact of AI overviews in order to prioritise content optimisations accordingly.
As Google might expand AI overviews to include high-cost queries when enabling ads, it’s important that all online retailers optimise their product pages taking into consideration not only traditional SEO practices but also the data requirements of AI overviews. These changes and ecommerce SEO strategies should then be monitored and updated based on data and trends.
Back in our March search and industry update blog, we covered the conclusion of the first Spam update of 2024. Alongside this update, Google also announced three new spam policies, one being the site reputation abuse policy which was due to come into effect in May.
Contrary to speculation, sites that were being hit by algorithmic penalties concerning the site reputation abuse policy, Google’s Search Liaison, Danny Sullivan, confirmed that the algorithmic actions targeting site reputation abuse have not yet been launched:
In addition to this, Sullivan also added that current manual actions only affect specific content and are not sitewide. However, he did not comment on the scope of the algorithmic actions once they have been rolled out.
In a recent article published by Glenn Gabe, he reflected on the March 2024 Google core update, highlighting the transition of Google’s helpful content system (HCS) into its core ranking system.
An overview of the previous helpful content system
Previously, the helpful content system was a separate classifier. This means that the old HCS classifier, evaluated websites at a site-wide level and could cause significant drops in rankings if a site was deemed to have “unhelpful content”. Since the release of the 2024 March Core Update, the HCS is now part of the core ranking system.
How the helpful content system has changed in light of the core update
Now embedded into the core ranking system, the HCS now works alongside other systems to provide an overall evaluation of a website’s helpfulness. With this integrated approach, it means that a site with generally positive scores from some systems could still see a poor overall performance if the content helpfulness system scores it negatively.
Advice for SEOs and website owners
Gabe wrapped up his article with one key piece of advice for SEOs and website owners, stating that if you experience a significant drop in positions, you must reevaluate your websites overall and not focus on individual pages. He hammers home the fact that we must no longer think of the helpful content system as a standalone system and need to remember that it is working in tandem with other systems to determine the helpfulness of content.
For the past few weeks, Google Search has been particularly volatile, particularly from May 17th through to early June. Google has not officially reported any updates during this period so the causes behind these fluctuations remain uncertain. Despite the lack of confirmed reports from Google, the SEO community and website owners saw a significant surge in SEO-related discussions, and many tracking tools continue to display high volatility, surpassing previous levels.
In possibly related news, SEO commentators like Lily Ray have observed a surprising trend: large, popular brands are seeing significant increases in visibility seemingly overnight. These brands have started ranking higher for non-branded keywords, despite having much lower or nonexistent rankings just days before.
Robin Fry, from Properllernet, explored this further, investigating the potential causes as to why Google is favouring big brands more than it has ever done before. The analysis suggests that these ranking increases are affecting service and blog pages rather than homepages. Even without content changes or relevant content on the ranked pages, these brands are surpassing other sites, clearly benefiting from their stronger brand names.
Fry concludes that factors such as links, brand mentions, or user behaviour might play a role, or perhaps RankBrain or another AI element of the algorithm has improved in associating certain brands with specific keywords. However, this spike in visibility raises more questions than it answers.
Keep up-to-date with our dedicated algorithm and search industry round-ups. For any further information about these posts, or to learn how we can support you, get in touch today.
