Twitter is “tanking” amid Threads’ surging popularity, analysts say

Data shows that a sudden spike in interest in Meta’s Threads—which surpassed 100 million sign-ups in five days, Mark Zuckerberg boasted yesterday—has likely already put a tiny dent in Twitter’s traffic, The Wall Street Journal reported.

The news comes after a tweet from Cloudflare CEO Matthew Prince went viral. In it, Prince shared a Cloudflare chart showing that since January, Twitter traffic—compared to other popular websites—has been “tanking.”

During Threads’ first two days online, Twitter traffic dropped by 5 percent compared to the same two days in the prior week, web analytics firm SimilarWeb reported. When measuring year over year, Twitter’s traffic dropped by 11 percent.

“We’ve been reporting for a while that Twitter is down compared with last year,” David Carr, a SimilarWeb senior insights manager, told CNN. “But Threads seems to be taking a bigger bite out of it.”

While Threads continued growing, Twitter execs Elon Musk and Linda Yaccarino started sharing a new metric for user engagement that they seemingly found more encouraging than the somewhat dismal traffic reports.

“Don’t want to leave you hanging by a thread… but Twitter, you really outdid yourselves!” Yaccarino tweeted. “Last week we had our largest usage day since February.”

Musk responded to that tweet, clarifying what Yaccarino meant by “usage.”

“Cumulative user-seconds per day of phone screentime, as reported by iOS & Android, is hardest to game,” Musk tweeted. “I think we may hit an all-time record this week.”

While many Twitter users debated whether this new metric was meaningful—and some claimed it wasn’t even possible to track—the T(w)itter Daily account explained “why Elon changed Twitter’s metrics” from tracking monthly daily active users (mDAU) to user-seconds per day.

“The old mDAU metric” included bots and “people who got a Twitter notification on their phone but didn’t open the app,” while the “new metric is much harder to manipulate,” T(w)itter Daily tweeted.

T(w)itter Daily also claimed that Twitter 2.0 data scientists had verified that “despite some claims to the contrary, it absolutely is possible for Twitter to track how much time people use the app on iOS and Android.”

Musk responded to T(w)itter Daily’s thread, confirming that bots can still manipulate the new metric “via humans with lots of phones, but that is [more than] 100 times more expensive than bots.”

Right now, CNN reported that Threads is “on pace to rapidly pass Twitter’s audience size.” Both SimilarWeb and Cloudflare told CNN that Twitter traffic trending downward appears to have been accelerated by Threads’ launch, which seems to pose an obvious risk to Twitter’s business. It’s possible that Yaccarino’s and Musk’s tweets have introduced a new metric in an effort to reassure advertisers that Twitter engagement has gone up, despite reports of traffic tanking.

T(w)itter Daily went so far as to claim that the reason that Twitter user activity is up despite Threads launching—and despite its own decision to start rate-limiting tweets—is because “all publicity is good publicity.” Following that logic, all reports comparing Threads’ success to Twitter’s decline, including this one (which links to multiple tweets), ultimately lead Twitter users back to Twitter.

Meta and Twitter did not immediately respond to Ars’ request to comment.

Twitter restricting Threads links?

While Musk and Yaccarino have seemed to laugh off Threads as a non-threat to Twitter’s business, users have noticed that Twitter seems to be acting defensively by seemingly restricting users from discovering Threads links on the platform.

The Verge pointed out that searching on Twitter for “threads.net” returns a chaotic list of users who have their Threads account in their display names, but no actual links to Threads. Narrowing the search by using the “url:” search operator to return results only at the “threads.net” URL—using the search term “url:threads.net”—currently returns zero results, Ars confirmed.

Twitter has not confirmed that Threads links are being intentionally restricted from search. It seems possible that’s the case, though. Previously Musk temporarily restricted Substack links on the platform out of fears that Substack’s new app Notes was too similar to Twitter to be promoted on Twitter. And recently, Musk has threatened to sue Meta, calling Threads a copycat.

Users who want to discover Threads links on Twitter can use a workaround, The Verge discovered. Relevant search results appear by tweaking the search operator formatting and searching for “url:“threads net”” instead of “url:threads.net.” Twitter users can also still search for individual Threads posts by using the link to any Threads post as a search term.

In addition to seemingly limiting searches for Threads links, Twitter also appears to be restricting Threads from trending on Twitter, a Meta engineer, Philip Fung, posted on Threads.

“Seems like Threads is censored from Trending,” Fung said, sharing a screencap of Twitter’s trending topics and pointing out that it was “unlikely” that Threads “is not in the top 30 US trends for the last day.”

“Makes sense, but also shows how hard it is to stick to your principles when it is against your interests,” Fung said, seemingly jabbing at Musk for censoring Threads when Musk is normally known as the anti-censorship guy.

Zuckerberg doesn’t seem to mind that Twitter is potentially limiting Threads from trending, responding to Fung’s tweet by joking, “Concerning” with a cry-laughing emoji. Instagram CEO Adam Mosseri responded a little more earnestly, posting, “It is what it is. More reason for us to focus on making Threads better and fast.”

Meta’s founder has insisted that Threads’ explosive growth—instantly taking the crown from ChatGPT as the fastest-growing consumer app—has been “mostly organic.”

“We haven’t even turned on many promotions yet,” Zuckerberg posted on Threads.

Once Meta begins promoting the app, Twitter’s traffic could take an even bigger hit, Zuckerberg seems to suggest.

Is Threads a threat to Twitter?

It’s still unclear whether all those people who signed up for Threads will actually end up becoming engaged users.

Two days after Threads launched, The Verge reported that internal data showed its approximately 30 million users had collectively posted more than 95 million times and shared 190 million likes. It’s a modest level of engagement, but not Twitter levels of engagement. Twitter users tweeted 500 million times a day in 2022, according to data gathered by marketing consultant David Sayce.

Mosseri has said that Threads’ “goal isn’t to replace Twitter,” but “to create a public square for communities on Instagram that never really embraced Twitter and for communities on Twitter (and other platforms) that are interested in a less angry place for conversations, but not all of Twitter.”

Meta is seemingly aiming to attract only a portion of Twitter’s audience while creating a more advertiser-friendly version of Twitter—by turning Threads into a platform that Mosseri said will not “encourage” divisive political and news posts.

Seemingly in at least Mosseri’s view, Twitter and Threads will likely always coexist and battle each other (and other platforms) in bids to increase user engagement and attract advertising dollars. One side seemingly gambling on suppressing divisive content, and the other seemingly gambling on not suppressing divisive content.

That version of the future makes it seem like it will be up to users to decide which platform suits their needs or if they need both platforms for different reasons. For some users dividing time between platforms, it seems like Twitter could remain a source for news, while Threads could become a preferred platform for hosting easily discoverable deep discussions.

The director of the Stanford Internet Observatory, Alex Stamos, performed an “unscientific test” to see whether his own posts got more engagement on Twitter or Threads by measuring likes and replies over a 23-hour period. He shared identical posts on each platform and “saw significantly more engagement on Threads”—where he has fewer than 10,000 followers—“than on Twitter”—where he has more than 100,000 followers.