Twitter traffic is ‘tanking’ as Meta’s Threads hits 100M users

cnbc.com

8 points·jb1991··5 comments

"[... Twitter traffic] down 5% for the first two full days Threads was generally available compared with the previous week. [...]down 11% compared with the same days in 2022."

The 11% is way less than I expected. I don't think the July/4th week is a good baseline for the 5% comparison.

Social networks don't die quickly. But all the trends are down, consistently, persistently, and seemingly irreversibly, and Twitter's traffic can go down 5-11% only so many times, before it no longer matters.

CloudFlare is only showing data from DNS queries to their 1.1.1.1 service. Twitter also happened to introduce rate limits at the same time Threads was released. While the article isn't saying it directly, it seems to be implying that Threads is causing this dip in requests, when in reality the majority of Twitter users are now making fewer requests to the website due to rate limits.

Another point in the article suggests also that the large number of Threads users were never Twitter users in the first place. So how could Threads have caused such a massive drop in Twitter DNS queries?

> 'Threads is being touted by Meta executives as a more positive “public square” for communities “that never really embraced Twitter.”'

> when in reality the majority of Twitter users are now making fewer requests to the website due to rate limits.

DNS data wouldn't show this; in general, a client will make _one_ DNS request for a given domain, then not bother again until the TTL expires (in Twitter's case, 544 seconds). Or, er, not until whenever it feels like; some clients don't honour DNS TTLs.

Stop thinking critically. We’re all trying to channel our Two Minutes Hate for Musk and we don’t like distractions.

Reply on news.ycombinator.com