On Monday, May 27, news broke that the closely guarded secrets of Google’s search algorithm had been leaked from inside their search division.
Rand Fishkin, Co-Founder of SparkToro and Snackbar Studio, revealed that he was emailed by a credible, anonymous source who ultimately shared over 2,500 pages of Google internal files with him.
A sceptical Fishkin reached out to a few friends who were former Google employees, and two of the three validated that the leak was “most likely legitimate”.
In a blog article detailing the leak, Fishkin outlines his analysis of and findings from the data, which he analysed with Founder and CEO of iPullRank, Mike King.
An Authentic Leak
When news broke, the source wished to remain anonymous. However, the following day, they chose to reveal their identity in a video uploaded to YouTube.
And on 29 May, Google confirmed that the leak is authentic.
The source, Erfan Azimi, is the CEO and Strategist of digital marketing agency EA Eagle Digital.
He wished for Fishkin to publish a detailed article on the leak and exposed some of the “lies” being spread through Google’s public statements about how their algorithm works.
Azimi’s motivations appear to align with those of Fishkin:
“To hold Google accountable for public statements that conflict with private conversations and leaked documentation, and to bring greater transparency to the field of search marketing.”
Interestingly, King’s take on it is that it’s more likely Google has been “gaslighting” us rather than intentionally lying, to deceive spammers and throw SEOs “off the scent of how to impact search results.”
What Have We Learned So Far?
The data being shared is most likely dated around August 2023, so how closely aligned it is with the current algorithm is up for debate. I imagine there isn’t a huge difference, but we can see the disparity with the absence of AI Overviews in the data.
Fishkin urges us to view the data not as “proof that Google uses XYZ in their rankings”, but rather as a “strong indication” of how the algorithm works.
I think it’s also important to note what King alluded to - that we don’t know if Google is using everything in the data.
People will be analysing the data for years, but Fishkin has listed five discoveries. Some provide greater clarity on how we assume Google operates, and others expose public statements as “lies.”
1) Types of Clicks
Google has consistently denied that click data is not used in their algorithm, but this isn’t true.
The data suggests that Google can filter out clicks depending on whether they want them to have an impact on their ranking systems.
We can determine this by references in the documents to “goodClicks, badClicks, lastLongestClicks, impressions, squashed, unsquashed, and unicorn clicks.”
Google also appears to measure the length of clicks (see our article on pogo-sticking), and impressions.
2) Google Chrome
Google’s browser (Chrome) is used to access the full clickstream of its users, something which they have previously denied in public statements.
The data from Chrome allows them to track multiple metrics that can be obtained from views to individual pages and entire domains.
This data, for example, appears to be used by Google when it creates Sitelinks, ranking your individual pages by the most-to-fewest clicks.
It seems that people who search for Atlas SEO actively want to know more about us than visit our Blog and Services pages.
If we look at user intent, the number of clicks on this page could indicate that searchers want to know more about our brand (more on this later).
But, hey, at least we’re getting interest in our fantastic content creation services!
3) Whitelisting Domains
The documents contained a module on “Good Quality Travel Sites”, which suggests that Google is whitelisting domains for travel queries.
This, combined with similar references in the data to COVID-19 and elections, indicates that Google is prioritising websites for queries that are related to controversial or problematic queries.
I think most of us would welcome this prioritisation, but it’s likely to be a pain for businesses in the travel sector which could be facing an uphill struggle to rank highly. And who knows, there may be other whitelisted sectors that are yet to be discovered.
4) Quality Rater Feedback
Google contracts quality raters to provide feedback on search results and provide suggestions on how to improve them.
According to Fishkin, there is evidence that some of that feedback has been directly incorporated in the search systems, but he was unable to determine how influential they were, or what they are used for.
But a point of interest here is the advice for marketers not to “dismiss how important it is that quality raters perceive and rate their websites well.”
It’s good practice to make sure your website is not just SEO optimised but created and maintained in a way that prioritises the user experience.
5) Classifying Links by Clicks
Google classifies its link indexes in three tiers—low, medium, and high quality—and uses clicks to determine which one they belong in.
A link with no clicks, for example, will fall into the low-quality index. Google simply ignores these links, so they will not affect your website’s ranking.
A link with a high number of clicks, however, is categorised as high quality and will pass on ranking signals. And once this type of link is “trusted”, it can transfer PageRank and anchors, or be filtered/demoted by link spam systems.
This ties in with what we see in our link building projects. Over the past five to ten years there’s certainly been a shift in requests for websites with high traffic, and this is now an established practice.
A website with high traffic is more likely to have greater user engagement, which, in turn, is likely to generate the clicks needed to place a link in the high-quality index.
Lessons for SEOs and Marketers
The level of technical analysis needed to understand the data is not for the average layperson. Those with the knowledge to analyse the data will relay their findings of what Google considers important; informing SEOs and marketers what to focus on, and what to ignore.
From Fishkin and King’s analysis, there are five takeaways for SEOs and marketers:
1) Prioritise Your Brand
Google prioritises established brands in the search results over small businesses and independent sites. And if you’re contending with whitelist sectors like travel, then your goal for greater organic traffic will be difficult.
If you’re a small fish in a big pond, the advice is to focus on building a reputable brand outside of Google search to improve your organic traffic and ranking.
2) E-E-A-T Questioned
There are question marks over the influence of E-E-A-T as a significant ranking factor. Fishkin suspects that its elements are likely tied up with what Google considers important in the ranking systems, so it’s hard to determine its efficacy as an SEO strategy.
That’s not to say that E-E-A-T strategies should be abandoned altogether. There’s a good chance that it still carries ranking benefits.
Fishkin says he’s worried it’s “80% propaganda, 20% substance” but I disagree. If the elements of E-E-A-T are tied to Google’s ranking factors, then it should still serve as good SEO practice.
3) User Intention is Important
A good marketer will know that the user should always be the priority when creating an effective strategy, and user intent is also at the core of a solid set of keywords.
The leak reveals that user intent can sometimes be a greater ranking factor than content and links.
Again, this boils down to clicks. If a user scrolls past X number of links in the SERPs and clicks on a link that is more relevant to their search query, that action tells Google to rank the page higher than those above it.
Of course, this makes sense. Google wants to position more clickable links as high up as possible, even if the links above it have excellent on-and-off page signals.
Does this mean you should give up on your content and link building strategies? Absolutely not. They are still two reliable SEO methods that will improve your site’s authority and ranking.
The takeaway here is to make sure that your content is created with user intent in mind, and link building will still get you the clicks needed for your site to become trusted in the eyes of Google.
4) PageRank Isn’t Dead
Many in the industry assumed that PageRank was a dying metric, but the leaked documents suggest that it still holds importance in search index and rankings.
So, while clicks are important for placing a link in the high-quality index, the number and quality of your backlinks are still important metrics for ranking.
Unsurprisingly, Page Titles are still an important feature that will help your ranking. Less importance, it seems, is given to the weight of anchor text links.
5) SMEs and Newer Publishers
According to Fishkin, SMEs and publishers with young domains should prioritise establishing credibility, navigational demand and a strong reputation before investing in SEO.
As mentioned above, prioritise building your brand because Google will favour established and popular brands in the search results. Only when your brand is well established can you start to think about competing with the big players.
This is not to say that you should give up on SEO. Following best practices will help Google understand your website better and therefore rank it higher. They can also help you to establish your brand and build its reputation.
But the warning here is around expectations. As a smaller business or newer publisher, it may not give you the ROI you’re looking for until you’re more established.
What Does the Future Hold?
Even in the early days of this leak, there’s a lot to consider and it’s only going to get more interesting as the data is further analysed and conclusions revealed in the coming weeks, months and years.
It’s important to note that this is a screenshot of an algorithm that constantly evolves. While there is nothing earth shattering in the early analysis, I think the data reveals some SEO fundamentals that should be considered in any business's marketing strategy.
From Fishkin and King’s early analysis, we can ascertain a few factors to consider in our SEO strategies moving forward.
Anyone who is new to SEO and/or owns a website that is relatively young (compared to established brands) should focus on building a brand with a strong reputation. Contrary to Fishkin’s advice, a good digital marketing agency offering online reputation management can offer guidance and services to achieve this goal.
The leak is aimed at exposing the lies behind some of Google’s public statements, but we can probably trust their claim that the user is ultimately the most important factor in ranking. Logically, that fits perfectly with Google’s mission statement.
Our Advice
There are some basic SEO practices you can follow to help improve your ranking.
-
Design your website with user experience at the fore.
-
Produce high-quality, human-written content that answers your audience’s queries.
-
Build a website that is easy to navigate.
-
Make sure your website is technically sound, with accurate and concise metadata that allows Google to understand what your website is about.
We can see from the leak that Google is using clicks to gather data and rank websites, so these practices will help you to generate the types of clicks you need to improve your positions in the SERPs.
If you are unsure how best to implement them, conduct comprehensive A/B testing to determine what works best for your site.
While SEO is an “it depends” game, this approach will give your website the best chance possible of improving these metrics:
-
User engagement
-
Organic traffic
-
Organic backlinks
-
Brand trust and authority
For now, keep an open mind about the information in this leak before making any drastic changes to your SEO strategy, and let’s see how it all unfolds.
