Traffic may be artificially increased by X’s new link experiment on iOS. Following the change, websites like Substack and Bluesky saw a dramatic rise in “fake” views. Nick Eubanks, vice president of owned media at the digital marketing platform Semrush, attributes this to a new behaviour that preloads information before users click on it.
Eubanks tells a news agency that, “this is a classic example of metrics distortion brought on by product experimentation at the platform layer.” The new experiment allows you to interact with the like, repost, reply, and bookmark buttons while viewing the webpage by having X collapse a post when you click on its link. In the past, engagement with the original X content was impacted when X’s in-app browser clicked a link in a post and the loaded page completely blocked the X post.
According to Eubanks, “X’s new browser is pre-loading link content in the background, which means the system fetches the destination page before a human actually taps or views it.” This “inflates analytics in a few key ways,” he continues, including by making click rates appear higher and deceiving publishers, advertisers, and artists into believing they are getting more traffic when it may not “correlate to real human visits.”
Related “Views” are lies on TikTok, YouTube, X, and everywhere else. Chris Best, CEO of Substack, was first thrilled by the increase in visitors to his website after X’s update, but he later realised that “most of the apparent lift is fake.” Best claims that “even after correcting for the fake views,” Substack continued to see a rise in traffic.
The site’s analytics for tracking logged-out daily active users have been “ruined” by X’s new preloading technology, according to Bluesky product manager Paul Frazee. Frazee continues, “X has started opening links in the background to make them load faster… but it has caused a bunch of other sites to get extra traffic that appears real.”
The new link configuration, according to X product leader Nikita Bier, resolves a prevalent grievance from creators, who frequently discover that content with links receive less exposure on the platform. Bier further describes that “people forget to Like or Reply because the web browser covers the post,” Bier writes. “As a result, X is unable to discern whether the content is of high quality.”
Preloading may increase interaction on X, but it may eventually hurt publishers and creators outside of X by making it harder for them to identify the source of their traffic. The distinction between user involvement and machine behaviour will become more hazy as a result of metrics inflation brought forth by interface trickery, preloading, autoplay, and AI summarisation, according to Eubanks. “Platforms that wish to gain credibility with advertisers and creators must be open about how engagement is measured and keep previews apart from users.”
The wider ramifications go beyond specific publishers. “We’re entering an era where metrics inflation through interface tricks, preloading, autoplay, and AI summarisation will blur the line between user engagement and machine behaviour,” Eubanks states. Analytics firms are under pressure to create more advanced techniques for identifying real human contact as a result.
The circumstance poses a trust issue for advertising. How can companies reliably analyse campaign performance or allocate funds if platform modifications have the potential to unexpectedly inflate traffic numbers? The event serves as a reminder of the influence social media platforms have over the larger online community.
Despite years of cautions about creating audiences on rented land, the X scenario also highlights how reliant publishers have gotten on platform-driven traffic. Entire business models may appear to change on paper when a single platform modifies its technological implementation.
X’s preloading experiment demonstrates how platform modifications can affect the entire web ecosystem and skew the data that marketers and publishers depend on. Platforms that wish to gain trust with marketers and creators must be open about how interaction is measured, as Eubanks points out. Since this won’t be the last time a large platform’s technical choices have unforeseen repercussions for everyone else, the episode serves as a wake-up call for the industry to develop better techniques for separating real user behaviour from automated platform activity.
We use cookies to ensure that we give you the best experience on our website. If you continue to use this site we will assume that you are happy with it.
