Online spam is still a major problem in our daily lives, and as we all know, it is one of the major issues at the heart of Google’s work, which deploys tools and resources to make the search experience ‘safe and spam-free‘. As per tradition, the Mountain Views team shared the results of the work carried out on this front in the past year, anticipating some of the data of the Webspam Report 2021, but also presented the ‘official name’ of the algorithm delegated to anti-spam activity, which is called SpamBrain and has been at work since 2018.

Fighting spam, the work of Google SpamBrain

Introducing this special algorithm is an article by Cody Kwok, Principal Engineer for Google, who describes SpamBrain as an “artificial intelligence-based spam prevention system”.

Google presenta SpamBrain

First launched in 2018, Google has been steadily improving the performance of this algorithm, which also has the notable feature of being “a robust and evolving platform to address all types of abuse“.

With an increasing volume of sophisticated spam produced every day found on the web, Kwok says, SpamBrain’s ability to identify “disruptive and malicious behaviour among billions of web pages has allowed us to keep over 99% of searches free of spam” – an achievement that, in truth, was first claimed in the 2019 Webspam Report.

In more detail, during 2021 SpamBrain identified almost six times more spam sites than in 2020 (when it already claimed to have detected 40 billion spam pages every day), and this resulted in a significant reduction in hacked spam (-70%), which was one of the most commonly observed types of spam in 2020, and gibberish spam on hosting platforms (-75%). All in all, Google was able to detect 200 times more spam sites in 2021 than when it started its battle almost two decades ago.

SpamBrain’s activities and Google’s anti-spam efforts

The other fronts of Google’s battle concern the protection of search quality and user safety, and according to the report significant progress was also made in 2021 in these areas other than traditional web spam, and in particular in the fight against link spam, scams and online harassment.

In the first case, the link spam update launched in July 2021 is bearing fruit “to broadly identify unnatural links and prevent them from affecting search quality”, so as to protect links, which remain a key signal for Google in “discovering and ranking results in a meaningful way”.

Great attention was also paid to fraud, which remains a major threat to the safety of online users: expanding on the work of 2020, Google launched several algorithm updates that led to a 40 per cent reduction in fraudulent results, and the improved coverage ‘allowed us to protect people from many more types of scams, in addition to the customer service requests we’ve been fighting in recent years’.

Moreover, again to protect user safety, Google has extended SpamBrain’s action ‘to address online harassment and, for name queries, reduce the importance of sites with exploitative removal practices’ – and, thus, Kwok reveals that SpamBrain has been the algorithm at work against predatory sites since last June.

Reducing the effects of ranking manipulation

In addition to spam, Google continues to be ‘working hard to reduce low-quality content and ranking manipulations‘, combating behaviour that keeps right on the edge of quality guideline violations, but ‘is still manipulative in nature and degrades the user experience’.

For example, the article continues, one aspect of this initiative was to improve the ranking of product review queries, where content “often consisted of rewritten product descriptions and not first-hand, authentic reviews”. As we know, there were two substantial updates to the way the search engine evaluates product reviews in English in 2021 (and one in this 2022, just a few weeks ago), which resulted in a significant reduction in low-quality reviews, while promoting those that boasted better content and expertise.

Ultimately, Kwok recalls and reiterates that Google’s commitment is “to ensure that nothing stands in the way of people finding the most useful content through Search” and, thanks to SpamBrain, it seems that this goal is within reach. It’s true that some sites can still get away with some spam or black hat SEO tactics, but today’s news and Webspam Report numbers confirm that Google is constantly testing new methods to detect and prevent those sites (and many others) from ranking high in Google search using illegal and unfair techniques.