There’s been a lot of chatter over the past week about what’s happening to Twitter’s traffic and engagement, as well as that of Threads, and whether the two are related and ~what’s it all mean for the future of social media~
This is my attempt to cut through some of the noise, with actual data – and empirical observations.
All derives from this week’s column in the @WSJ
1/
https://www.wsj.com/articles/heres-how-twitter-could-become-irrelevant-b2b027af
Here’s the conclusion, for those who want the headline and then would rather move on with their lives.
Traffic to Twitter appears to be declining over time. Threads seems to be responsible for a little of that decline, when it first launched. But really, what’s happening is a story of long-term decline in engagement with Twitter. 2/
Another conclusion:
Much has been made of the surge and sudden drop in usage of Threads.
If you’ve never looked at the chart of the usage of a service when it goes supersonic, you might naively think this means Threads is already in trouble.
But a saw-tooth pattern where adoption is followed by a decline in usage – and then subsequent waves of re-engagement – is normal.
3/
Remember, this is just *week 1* plus 2 days for Threads. Fastest-growing app in history is probably going to do some weird things in the first months of its life.
And anyone who is a veteran of the endless cycles of "oh Mastodon saw a surge / oh now everyone is leaving" knows that this is how adoption of a service works.
As Eugene Wei told me, we’re going to know a lot more in a month, 6 months, a year. All conclusions at this point are premature.
4/
So now I want to point you to others who are being smart about this and who have explained Twitter’s long-term decline.
First, there’s this piece by @taylorlorenz which points out what should be blindingly obvious: Twitter’s real competition isn’t Threads, it’s TikTok.
https://www.washingtonpost.com/technology/2023/07/07/twitter-dead-musk-tiktok-public-square/
5/
And here’s a conclusion that @profgalloway and I both reached: Twitter’s problem isn’t threads, it’s Elon Musk’s leadership. By driving away users, mostly through changes to its algorithm and secondarily through his editorial decisions, he’s the real disruptor here.
“The misperception about disruption is that it’s a function of innovation or excellence. In fact, disruption is driven by stasis and the incompetence of incumbents.”
https://www.profgalloway.com/threadzilla/
6/
Last week, Twitter became MySpace: a social network void of innovation being slowly euthanized by Meta. In less than a week, Meta’s Threads registered 110 million users — equivalent to the combined population of Germany and Australia, and the most violent corporate disruption in platform history. What can be learned? Bloat The business strategy that […]
For those of you who stuck around this long, here’s a WSJ exclusive. No one else has this data. It comes from Diffbot, which crawls and indexes the web like Google does, and then sells access to all that data to companies.
Guess what effect all this decline has had on Twitter citations – links, embeds – in the news media, regardless of political orientation?
Zilch. Nada.
This is one area Twitter retains its death-grip on attention.
This is a long thread already, but it’s just a portion of what I was able to dig up or convey on what’s really happening with Twitter and Threads, Musk and Zuckerberg and the rest. Many insights are not my own but belong to @Noupside
Eugene Wei and Maelle Gavet, head of Techstars
I wish I could have quoted them at length. What I learned from talking to them this week is that we are in for VERY interesting times for social media.
fin/
@mimsical this is really interesting! twitter had become so commonly cited in the new media (because so many journalists are on twitter? because politicians and celebs use twitter as the medium and sidestep other media so they have no choice?) its a bit of a self-reinforcing pattern, the media cites twitter which makes it important, which means the vip users still want to be seen on it. but it also doesn't seem like a very monetizable asset for twitter.
iirc twitters income statement in the 10ks and 10qs before musk were pretty solid. its really a self-inflicted decline.
@mimsical What that graph says to me is Twitter's becoming a right-wing outfit. The blue line's in a slow but steady decline, while the red line, while volatile, is increasing.
My take on this is Musk bought Twitter to control the political narrative, and it looks like he's succeeding.
@mimsical seems like a good analysis for twitter/threads, as their value (to the business) depends on growth.
The comparison with fedi is hard to do correctly though. Referencing “everyone is leaving fedi” narratives is referencing bad journalism in ways you may not have meant. Fedi doesn’t need numbers to be a success, it already is a success. For those running many instances having more numbers just means a greater infrastructure and moderation burden. For many users there’s also diminishing returns from following more people. The value in growth for Fedi is diversity not quantity. Happy and safe users is what success looks like for much of Fedi, it’s Twitter where more users means more numbers to milk for profit.
Twitter user numbers means more resilience. The resilience number for Fedi is measured in servers, admins, and devs. Running a fedi server is getting easier, less technical, and outsource-able.
I suspect for many users the value of Fedi growing is:
* implying that fewer people are on networks they dislike
* having all the people they care about in one place
* more people developing UI/UX
—
I think it’d be much more interesting to do a user-centric comparison of the networks. Almost ignore growth, except where it relates to demographics.
Demographics are interesting to understand marginalisation and representation, the usual ones like age/race/gender/cultural/location/lgbtq+ but also disproportionally represented groups that have found a home, like furries and trans people, or survival bias from the initially complicated onboarding: technical-ability.
There are various groups/personas of users that it may be good to include in an analysis, for example:
1. friend groups
2. shitposters
3. meme/news consumers
4. those needing followers for their living
5. those marginalised offline/online
6. trolls/assholes/nazis
7. companies
I suspect current representation is something like:
* twitter now: 2,4,6,7 and some 3
* fedi: 1, 2, 4, mostly 5, and some 3 and 6
* twitter later: 6 or 0
I think fedi for the most part has no interest in 6 and 7, and will typically cut-ties with servers who have 6 in particular.
Fedi does seem to lack a diverse range of demographics, but it is improving, and Twitter is rapidly losing what they had. The loss almost everyone can agree on is the big melting pot of humanity that Twitter was. That was very important to many, and so there’s interest in the husk that held it.
I don’t think the interesting thing is where twitter’s decaying husk blows in the wind. It’s interesting what it was, and more interesting which other things may be able to recapture the best parts of that.
No matter what, let’s not add nazis back to the melting pot if we ever have it again.
@mimsical one word.
GOOD!
Christopher Mims
bitcoin psychopath
Blake Eskin
John Refior
Matthew Eric Bassett
Carl B. Latro
Hermey G
Christophe Jammet
Medea Vanamonde🏳️⚧️ ♀
Mark Loundy
🌈 Andrew ☄️
Dr M
Erin Conroy
Brennan Stehling
Craig Mcgee
Kevin Schwarz
Rooney’s old account