Scroll to read more

Are entering the end game for Twitter?

The latest rate limits fiasco has many once again eulogizing the app – but are we really on the precipice of death for the blue bird platform?

In reality, probably not, though Twitter does seem to be heading into another period of significant challenge, with users scrambling to alternative platforms – mostly as a hedge for what may become of Twitter – while Meta’s long-awaited rival app now has an official release date.

First off, on the current state. As we reported over the weekend, Twitter has now begun limiting how many tweets each user can view in the app within a given time frame, in an effort to halt what Elon Musk has described as ‘extreme levels of data scraping’ by third parties seeking to use tweet data for their own purposes.

Twitter has since increased these thresholds to:

  • Verified accounts – 10,000 posts
  • Unverified accounts – 1,000 posts
  • New unverified accounts – 500 posts

Those upper limits are probably harder to reach than you might think for most regular Twitter users – which is both a positive and negative for Twitter itself. But the issue at play here is that Twitter makes virtually all of its income from ad exposure, which is reliant on people actually being exposed to ads in the app.

The fewer tweets Twitter can show users, the fewer ads it can sell, so from a fundamental standpoint, this actively limits Twitter’s opportunities for broader business success.

But Elon seems okay with it, at least for the time being.

I don’t know that his investment partners would share the same perspective. Though, I guess, if you get into the Elon business, you have to accept that you’re going along for the ride, whatever that may be.

The move also increases the difficulty level for new Twitter CEO Linda Yaccarino, who’s tasked with getting the company’s ad business back on track. Twitter has lost many ad partners in the 9 months since Elon took over at the app, effectively slashing its revenue intake by half, and while many of them have since returned (according to Musk) restricting their potential for reach isn’t likely to improve the platform’s appeal for ad spend.

Yet, at the same time, there is a logic to the seeming madness at play here.

As Elon notes, with the rise in businesses looking to tap into the generative AI wave, there’s an increasing need for conversational data to feed into these large language models (LLMs). Facebook already limits what non-users can access (which it made even more limited following the Cambridge Analytica scandal), as does LinkedIn, which leaves Twitter and Reddit as prime targets for data scraping.

That’s been exacerbated even further by both platforms recently upping the price of their API access. If businesses can’t afford the higher-priced API costs, they can still scrape platform data for free due to the open nature of the app. And that approach is what Twitter says has put increased strain on its servers, as more and more generative AI projects are launched.

Essentially, Twitter’s both not making money from these LLMs that are stealing tweet data, then onselling it via their own generative AI tools, while it’s also costing Twitter more money to service these increasing requests. Hence, the decision to implement rate limits, which Elon says are temporary, though it could be difficult to stamp out the process if those limits are to be lifted.

Add to this the fact that Twitter’s own system is seemingly overloading itself as a result of system errors, and Twitter clearly has some big challenges on its hands, with no easy solutions.

While at the same time, the platform’s other elements of concern for ad partners are also not looking so great.

Another point of contention for ad partners is brand safety, due to Elon Musk’s new ‘freedom of speech, not freedom of reach’ approach, which includes more lax rules around what people can tweet, with Twitter looking to reduce the exposure of rule-breaking content, as opposed to removing it.

The latest act on this front saw Elon declaring that ‘comedy is legal on this platform’ in support of a controversial episode of Theo Von’s podcast, in which Von’s interviewee, comedian Roseanne Barr, made seemingly anti-Semitic remarks.

The episode has been banned from other platforms, but Twitter has allowed it to remain, despite Barr’s questionable comments. And while Twitter has vowed not to amplify such (the comments clearly break Twitter’s own rules on Hateful Conduct), Musk himself retweeted the episode link, which has helped push it to over 60 million views in the app.

It’s actions like this that win Musk support within the ‘free speech’ camp, but raise more concerns among ad partners. And while Twitter has repeatedly claimed that exposure to hate speech is down in the app, and that it’s doing more to implement third-party ad verification, to reassure business users, it’s not the ideal setting for Yaccarino to be meeting with potential ad buyers to sell them on the opportunities of the app.

There have also been more questions raised about Twitter’s new approach to content moderation, and its increasing reliance on Community Notes to help police misinformation in the app.

According to research by Poynter Institute, while Community Notes does show promise in some applications, the actual systematic process that dictates which Notes are shown and which are not, based on filtering for political bias, means that 60% of the most-rated notes in the app are not publicly available.

The flaw here is that for a note to be shown, there needs to be agreement from both sides of the political spectrum that a note is necessary, and for some of the most divisive claims, that consensus is simply not going to happen. As a result, Poynter says that only around 8.5% of the Community Notes created are viewable by regular Twitter users.    

So while Community Notes has been beneficial in, say, highlighting an AI generated image, where there’s agreement that a note would be of value, in a broader context, it’s not helping to dispel misinformation, another concern for potential ad partners.

It is worth noting that Musk also recently clarified that Twitter still has over 4,000 moderators working for the company following its staff cull, so it’s not entirely reliant on Community Notes to weed out untruths. But the Community Notes approach, in combination with Twitter’s more relaxed rules around speech, doesn’t appear to be any kind of solution, at least not as the Twitter 2.0 team appears to have projected.

So, does that mean that Twitter’s on the outs, that people are done with Musk’s reformed approach, that advertisers will now turn their back on the platform?

Again, people have been refining their Twitter’s obituaries for nine months now, and seemingly against the odds, it’s still going, and it’s still a valuable, useful tool for breaking news and major events. But it is becoming less valuable for such over time, as Musk continues to try and push users towards his paid model, along with the varying degrees of trial and error to address these other concerns.

And with Meta now setting a date for the launch of its Twitter rival app ‘Threads’, it does seem that the door is ajar, and there could be an opportunity for a competitor to gain ground.

Indeed, downloads of Twitter rivals Mastodon, Spill, and Post have all surged in the days following the introduction of Twitter’s new rate limits.

I do think that these are just back-up plans for Twitter users who are being spooked by headlines of the app’s demise, and that this is not reflective of a broader shift away from Twitter, necessarily.

But Meta is investing a lot into its new app.

And with the scale of Instagram helping to boost take-up, maybe this is the start of the Twitter migration many have predicted.