The mapping platform Google Maps lives not only from purchased and self-recorded data but also from the users who make millions of contributions every day. Above all, the ratings and reviews are an important part, where quality plays a major role. Now the team gives interesting insights into how the rating system works and how fake reviews are blocked.
According to their own statements, Google Maps users submit millions of reviews every day, which are published on the map platform and also in web searches. But not all reviewers have good intentions or want to give valuable feedback, but of course, it is or would be a paradise for spammers. It is also becoming increasingly common for competitors to be intentionally badmouthed – knowing full well that many people look at these reviews in advance.
To filter this out, Google Maps has developed a three-stage system: Immediately after clicking Publish, a review is run through the algorithms to filter out obvious spam. If it goes through this step, the review can be published immediately.
But even after that, the algorithms check the rating for suspicious patterns and, if necessary, pass it on to the team, which consists of an unknown manpower size and is trained to “evaluate ratings” and unlock or block them accordingly.
The video attached below shows the whole process well and explains it in English. I think that shows very well how much effort goes into keeping the ratings of the card platform free of such things. And quite successfully, because I personally very rarely come across ratings or reviews that seem implausible or that are obviously spam.
Reviews are analyzed from three directions:
- < The content: Does it contain provocation, spam or off-topic content?
- The Reviewer: Has this Google account published questionable reviews in the past?
- The location: Was there any unusual activity, noticeable accumulation of 1 or 5-star reviews, or was the location in the media that suddenly became so interesting?
We are always working to keep the quality of the reviews high. This applies not only to the reviews but also to the published star ratings.
Well over a million are published every day, although unfortunately, the information does not indicate whether this refers to stars or texts. The first is much more common. In addition, it is not known whether the millions will be published or just submitted. Because the spam rate is likely to be quite high.