Reviews can be quite helpful online. But it becomes problematic when these are fake. Google has also had a problem with this so far. But an update to the algorithm should now help against fake reviews.

Whether address, opening times or contact options: Google now has some information about local companies. Reviews are also helpful for many interested parties, for example when they are looking for a restaurant near them.

In 2023 alone, Google processed around 20 million changes to information such as opening times or reviews every day and fed them into Maps. According to its own information, the US company was also able to “identify and remove more fake reviews than ever before” last year.

Google is taking action against fake reviews

In order to combat fake reviews better and faster, Google has tweaked its algorithm. A machine learning algorithm has been used for this since 2023.

This can examine signals daily over a longer period of time. For example, it is noticeable if a profile gives the same rating for several companies. The sudden increase in 1 or 5 star ratings for a company can also be recognized more quickly.

The use of the new algorithm has already shown success. Because in 2023, Google was able to detect and remove 45 percent more fake reviews.

Algorithm shuts down fraud network

With the help of the AI-based algorithm, Google was also able to shut down a network last year that specialized in fake reviews. The algorithm can “continuously analyze patterns” and, for example, recognize whether an account has previously submitted fake reviews.

Within a few weeks, Google was able to intercept five million fake review attempts in connection with this fraud network. All accounts in question have been deactivated by Google.

In total, thanks to the change in the technology used, Google was able to identify more than 170 million reviews that violated the guidelines in 2023. The US group was also able to remove more than 12 million fake company profiles.

Also interesting:


Leave a Reply

Your email address will not be published. Required fields are marked *